#Marr's computational model | Explore Tumblr posts and blogs

compneuropapers · 3 months ago

Text

Interesting Papers for Week 15, 2025

Surprise!—Clarifying the link between insight and prediction error. Becker, M., Wang, X., & Cabeza, R. (2024). Psychonomic Bulletin & Review, 31(6), 2714–2723.

Learning enhances behaviorally relevant representations in apical dendrites. Benezra, S. E., Patel, K. B., Perez Campos, C., Hillman, E. M., & Bruno, R. M. (2024). eLife, 13, e98349.3.

Symmetry breaking organizes the brain’s resting state manifold. Fousek, J., Rabuffo, G., Gudibanda, K., Sheheitli, H., Petkoski, S., & Jirsa, V. (2024). Scientific Reports, 14, 31970.

Stimulus-invariant aspects of the retinal code drive discriminability of natural scenes. Hoshal, B. D., Holmes, C. M., Bojanek, K., Salisbury, J. M., Berry, M. J., Marre, O., & Palmer, S. E. (2024). Proceedings of the National Academy of Sciences, 121(52), e2313676121.

Dynamic responses of striatal cholinergic interneurons control behavioral flexibility. Huang, Z., Chen, R., Ho, M., Xie, X., Gangal, H., Wang, X., & Wang, J. (2024). Science Advances, 10(51).

Bridging the gap between presynaptic hair cell function and neural sound encoding. Jaime Tobón, L. M., & Moser, T. (2024). eLife, 12, e93749.4.

Reducing the Influence of Time Pressure on Risky Choice. Jiang, Y., Huang, P., & Qian, X. (2024). Experimental Psychology, 71(4), 238–246.

Broadscale dampening of uncertainty adjustment in the aging brain. Kosciessa, J. Q., Mayr, U., Lindenberger, U., & Garrett, D. D. (2024). Nature Communications, 15, 10717.

Temporal context effects on suboptimal choice. McDevitt, M. A., Pisklak, J. M., Dunn, R. M., & Spetch, M. L. (2024). Psychonomic Bulletin & Review, 31(6), 2737–2745.

A computational model for angular velocity integration in a locust heading circuit. Pabst, K., Gkanias, E., Webb, B., Homberg, U., & Endres, D. (2024). PLOS Computational Biology, 20(12), e1012155.

A neuronal least-action principle for real-time learning in cortical circuits. Senn, W., Dold, D., Kungl, A. F., Ellenberger, B., Jordan, J., Bengio, Y., Sacramento, J., & Petrovici, M. A. (2024). eLife, 12, e89674.3.

Eye pupils mirror information divergence in approximate inference. Shirama, A., Nobukawa, S., & Sumiyoshi, T. (2024). Scientific Reports, 14, 30808.

Inferring context-dependent computations through linear approximations of prefrontal cortex dynamics. Soldado-Magraner, J., Mante, V., & Sahani, M. (2024). Science Advances, 10(51).

Noisy Retrieval of Experienced Probabilities Underlies Rational Judgment of Uncertain Multiple Events. Spiliopoulos, L., & Hertwig, R. (2024). Journal of Behavioral Decision Making, 37(5).

Evaluating hippocampal replay without a ground truth. Takigawa, M., Huelin Gorriz, M., Tirole, M., & Bendor, D. (2024). eLife, 13, e85635.

Future spinal reflex is embedded in primary motor cortex output. Umeda, T., Yokoyama, O., Suzuki, M., Kaneshige, M., Isa, T., & Nishimura, Y. (2024). Science Advances, 10(51).

The emergence of visual category representations in infants’ brains. Yan, X., Tung, S. S., Fascendini, B., Chen, Y. D., Norcia, A. M., & Grill-Spector, K. (2024). eLife, 13, e100260.3.

Cortisol awakening response prompts dynamic reconfiguration of brain networks in emotional and executive functioning. Zeng, Y., Xiong, B., Gao, H., Liu, C., Chen, C., Wu, J., & Qin, S. (2024). Proceedings of the National Academy of Sciences, 121(52), e2405850121.

The representation of abstract goals in working memory is supported by task-congruent neural geometry. Zhang, M., & Yu, Q. (2024). PLOS Biology, 22(12), e3002461.

Theta phase precession supports memory formation and retrieval of naturalistic experience in humans. Zheng, J., Yebra, M., Schjetnan, A. G. P., Patel, K., Katz, C. N., Kyzar, M., Mosher, C. P., Kalia, S. K., Chung, J. M., Reed, C. M., Valiante, T. A., Mamelak, A. N., Kreiman, G., & Rutishauser, U. (2024). Nature Human Behaviour, 8(12), 2423–2436.

#neuroscience #science #research #brain science #scientific publications #cognitive science #neurobiology #cognition #psychophysics #neurons #neural computation #neural networks #computational neuroscience

11 notes · View notes

jcmarchi · 2 years ago

Text

Using Language to Give Robots a Better Grasp of an Open-Ended World - Technology Org

New Post has been published on https://thedigitalinsider.com/using-language-to-give-robots-a-better-grasp-of-an-open-ended-world-technology-org/

Using Language to Give Robots a Better Grasp of an Open-Ended World - Technology Org

Imagine you’re visiting a friend abroad, and you look inside their fridge to see what would make for a great breakfast. Many of the items initially appear foreign to you, each encased in unfamiliar packaging and containers. Despite these visual distinctions, you begin to understand what each one is used for and pick them up as needed.

Robotic image recognition – illustrative photo. Image credit: MIT CSAIL

F3RM can interpret open-ended language prompts from humans, making the method helpful in real-world environments that contain thousands of objects, like warehouses and households.

F3RM allows robots to interpret open-ended text prompts using natural language, helping the machines manipulate objects. As a result, the machines can understand less specific human requests and still complete the desired task.

For example, if a user asks the robot to “pick up a tall mug,” the robot can locate and grab the item that best fits that description.

“We really want to figure out how to do that, so with this project, we try to push for an aggressive level of generalization, from just three or four objects to anything we find in MIT’s Stata Center. We wanted to learn how to make robots as flexible as ourselves, since we can grasp and place objects even though we’ve never seen them before.”

Robots learning “what’s where by looking”

The method could assist robots with picking items in large fulfillment centers with inevitable clutter and unpredictability. In these warehouses, robots are often given a description of the inventory that they’re required to identify. The robots must match the text provided to an object, regardless of variations in packaging, so that customers’ orders are shipped correctly.

With F3RM’s advanced spatial and semantic perception abilities, a robot could become more effective at locating an object, placing it in a bin, and then sending it along for packaging. Ultimately, this would help factory workers ship customers’ orders more efficiently.

“But before we scale up this work further, we want to first make this system work really fast. This way, we can use this type of representation for more dynamic robotic control tasks, hopefully in real-time, so that robots that handle more dynamic tasks can use it for perception.”

The MIT team notes that F3RM’s ability to understand different scenes could make it useful in urban and household environments. For example, the approach could help personalized robots identify and pick up specific items. The system aids robots in grasping their surroundings — both physically and perceptively.

“David Marr defined visual perception as the problem of knowing ‘what is where by looking,’” says senior author Phillip Isola, MIT associate professor of electrical engineering and computer science and CSAIL principal investigator.

“Recent foundation models have gotten really good at knowing what they are looking at; they can recognize thousands of object categories and provide detailed text descriptions of images. At the same time, radiance fields have gotten really good at representing where stuff is in a scene. The combination of these two approaches can create a representation of what is where in 3D, and what our work shows is that this combination is especially useful for robotic tasks, which require manipulating objects in 3D.”

Creating a “digital twin”

F3RM begins to understand its surroundings by taking pictures on a selfie stick. The mounted camera snaps 50 images at different poses, enabling it to build a neural radiance field (NeRF), a deep learning method that takes 2D images to construct a 3D scene. This collage of RGB photos creates a “digital twin” of its surroundings in the form of a 360-degree representation of what’s nearby.

In addition to a highly detailed neural radiance field, F3RM also builds a feature field to augment geometry with semantic information. The system uses CLIP, a vision foundation model trained on hundreds of millions of images to efficiently learn visual concepts. By reconstructing the 2D CLIP features for the images taken by the selfie stick, F3RM effectively lifts the 2D features into a 3D representation.

Keeping things open-ended

Each potential option is scored based on its relevance to the prompt, similarity to the demonstrations the robot has been trained on, and if it causes any collisions. The highest-scored grasp is then chosen and executed.

To demonstrate the system’s ability to interpret open-ended requests from humans, the researchers prompted the robot to pick up Baymax, a character from Disney’s “Big Hero 6.” While F3RM had never been directly trained to pick up a toy of the cartoon superhero, the robot used its spatial awareness and vision-language features from the foundation models to decide which object to grasp and how to pick it up.

If the bot sees two glass mugs and one of them is filled with coffee and the other with juice, the user can ask for the “glass mug with coffee.” The foundation model features embedded within the feature field enable this level of open-ended understanding.

“F3RM combines geometric understanding with semantics from foundation models trained on internet-scale data to enable this level of aggressive generalization from just a small number of demonstrations.”

Written by Alex Shipps

Source: Massachusetts Institute of Technology

You can offer your link to a page which is relevant to the topic of this post.

0 notes

perfectlytinyworkspace · 4 years ago

Text

Mitchell: The World as Exhibition

"The four Egyptians spent several days in the French capital, climbing twice the height (they were told) of the Great Pyramid in Alexandre Eiffel's new tower..."

"The Egyptian exhibit had been built by the French to represent a street of medieval Cairo...'It was intended,' one of the Egyptians wrote, 'to resemble the old aspect of Cairo.' So carefully was this done, he noted, that 'even the paint on the buildings was made dirty.'"

To the West, the rest of the world is archaic, antiquated spectacle – a curious view of the past, an antithesis to Western progress "in the right direction"

"The Egyptian visitors were disgusted by all this...Their final embarrassment had been to enter the door of the mosque and discover that...it had been erected as what the Europeans called a façade."

"Together with other non-European delegates, the Egyptians were received with hospitality–and a great curiosity...they found themselves something of an exhibit."

"What [can] this process of exhibiting tell us about the modern West[?]"

modernity as defined by the West positions its own culture as the norm and the mundane, from which "the ordering up of the world itself as an endless exhibition" can be procured as entertainment, curiosity, and an object of "interested study, intellectual analysis"

how can Islam's relationship with modernity be a positive one, defined as it is by this sort of cheapened commodification of its premises as entertainment for the "civilized West"?

Yuval Noah Harari in Sapiens posits that one of the key reasons for Western/European dominance is 'curiosity' and 'the cultural backing to question tradition/history.' Does the "curiosity of the European [encountered] in almost every subsequent Middle Eastern account" following the first Arabic description of 1800s Europe then become a defining factor/formidable strong suit of Western modernity?

Does the "[demonstration of] the history of human labor by means of 'objects and things themselves'" espoused by the European exhibitions of the latter 1800s speak to the beginnings of an increased focus on the "hardware" and "observable, empirical data" of any one subject of study?

Marr's 1982 argument in computational social science: any informational system can be analyzed via 1) problem/computation 2) algorithm 3) physical system hardware; science and especially neuroscience today emphasizes (3) and thus comes away, as argued by some, as an incomplete understanding of the way the brain works

i.e., a bird's mechanisms and purposes for flight cannot be deduced from the study of its physical hardware components alone (feather)

in the same way, an "object lesson" is not the definitive experience of another culture/lifestyle as imagined by the "Histoire du Travail" display of the 1889 Exhibition but rather an overemphasis on the "physical hardware" and material of that culture and European imaginations and (mis)interpretations of its computational/algorithmic functions and purposes

One major Arabic response to Western creations of spectacle is heavy documentation, "[devoting] hundreds of pages to describing the peculiar order and technique of these events–the curious crowds of spectators, the scholarly exhibit and the model...the systems of classifications...the lectures, the plans and the guide books–in short the entire machinery of what we think of as representation."

"They were taken to the theater, a place where Europeans represented their history to themselves..." – non-Western cultures are represented in the same fashion, as spectacle, as traditions and culture from the West's own history. In unifying their systems of representation the West has relegated the rest of the world, including the Arab world and Islam, to the same status as its own history

Is a society's/culture's own history also a form of subjugated knowledge? i.e., when history is remembered as tragic and painted as a negative, primitive state rather than a series of traditions and stories to be revered, does that indicate the modern knowledge vanquishing/subjugating the past knowledge?

"The Europe in Arabic accounts was a place of spectacle and visual arrangement, of the organization of everything, and everything organized to represent...some larger meaning."

"intizam al-manzar, the organization of the view" – is this...the panopticon? Western visual organization of spaces, spectacles, and symbols to convey their interpretations of non-Western cultures feels very much akin to a disciplinary technique to exercise power subtly over the bodies and <souls> of non-Western individuals/populations.

the spectator role holds no power – the spectator can only witness and be complicit to his own objectification

'objectification' in the sense that the individual becomes a material representation of 'the Orient' or 'the East' and all of his actions, thoughts, speech, mannerisms are subsequently first filtered through the lens of this representation to fit to the Western idea of 'his culture and people' before they are attributed to him, and the resulting communication of his identity and actions is so garbled and perverted that it really only serves to reinforce the West's perception of them.

"First, there was the apparent realism of the representation. The model or display always seemed to stand in perfect correspondence to the external world...Second, the model, however realistic, always remained distinguishable from the reality it claimed to represent...the medieval Egyptian street at the Paris Exhibition remained only a Parisian copy of the Oriental original."

Is this an example of Baudrillard's hyperreality? And if it is, does that mean that, as he states, 'neither the representation nor the real remains, just the hyperreal'?

furthermore, if only the hyperreal remains, what is the hyperreal? we know that the representation is the Parisian perception of the non-Western world, and that the real is the non-Western world itself (but is that world in the past or the present?)

so in this instance I suppose neither of them remain and the strangely perfect-but-not "Parisian copy of the Oriental original" is the only thing present – the "effect called the real world"

"...the world of representation is being admired for its dazzling order, yet the suspicion remains that all this reality is only an effect."

Is the "search for a pictorial certainty of representation" unique to the West?

was the creation of hyperreality uniquely borne of Western society? The argument is that the East is "...a world where, unlike the West, such 'objectivity' was not yet build in"

does Western hyperreality alone fall between Islam and an understanding reached with Western Judeo-Christian societies?

can the obstacle of exhibition/spectacle be overcome? Who needs to make the first step to overcome it, the West or the East? What does this first step look like, and it is actually possible to achieve given the nature of media coverage and social media in the modern era that creates a new hyperreality with regards to our understanding of the outside world?

The only objects and locations of value to Western modernity are those whose "pictorial certainties of representation" can inspire awe, wonder, marvel – or fit in with the overall Western representation of non-Western cultures as an exhibition of interest.

"The ability to see without being seen confirmed one's separation from the world, and constituted, at the same time, a position of power." – the panopticon guard!!

"To establish the objectness of the Orient, as a picture-reality containing no sign of the increasingly pervasive European presence required that the presence itself, ideally, become invisible." – this harkens to covert US support of Middle Eastern regimes that benefitted its own oil interests in the area while at the same time performing espousal of the area's "democratic rule" and "self-governance," "autonomy"

in pursuing an "authentic experience" as an outsider, the European spectator necessarily creates his flawed, hyperrealistic representation of the non-Western individual by appropriating "the dress and [feigning] the religious belief of the local Muslim inhabitants" as his disguise of invisibility and non-perception, despite "...being a person who had no right to intrude among them."

"Unaware that the Orient has not been arranged as an exhibition, the visitor nevertheless attempts to carry out the characteristic cognitive maneuver of the modern subject, separating himself from an object-world and observing it from a position that is invisible and set apart."

Western modernity is the European pursuit and implementation of a hyperreal representation of Islam with the underlying desire to both observe the true nature of the spectacle within and remain unobserved (thus holding onto a position of power)

"This, then, was the contradiction of Orientalism. Europeans brought to the Middle East the cognitive habits of the world-as-exhibition, and tried to grasp the Orient as a picture."

I'm not sure if this is still the prevalent world-view, or if it is one example of the Western tendency to impose its own perspective of the Middle East onto the reality of issues in Arab and non-Western countries

i.e., the viewpoint that democratization is the best and inevitable system of governance and steps away from it are 'regression' towards 'the archaic past'

when the Orient that is not created as an exhibition fails to meet the European spectator's expectations, that reality is 1) dismissed as inferior or 2) painted as corrupted by non-Western modernity, which is always 'straying away from the true form of the Orient,' which can now only be found in European definitions of the non-Western world in the eyes of the European tourist/scholar

very similar to Christian scholars stating that they 'saved Buddhism' by 'rediscovering the pure, original form of the religion' beneath 'idolatry and other corruptions of the core Buddhism'

#modern #notes

14 notes · View notes

luxe-pauvre · 4 years ago

Quote

Perhaps the recognition of his mortality gave [David] Marr's prose a grander perspective than the details of a model of vision might otherwise have warranted. He put his view of how the brain worked into a far broader, ethical context, telling us something about how we have evolved and the origin of our most deeply-held views in the effects of natural selection: 'To say the brain is a computer is correct but misleading. It's really a highly specialised information-processing device - or rather, a whole lot of them. Viewing our brains as information-processing devices is not demeaning and does not negate human values. If anything, it tends to support them and may in the end help us to understand what from an information-processing view human values actually are, why they have selective value and how they are knitted into the capacity for social mores and organisation with which our genes have endowed us.' It has been said that the density of Marr's mathematics means that his work has more often been cited than understood. This quip reveals that Marr's significance lies not in the precise detail of his computational models of vision - even his most ardent supporters accept that much of his book is now largely of historical interest - but rather in his overall approach. Unlike [Horace] Barlow, Marr did not think that the activity of single neurons could explain how circuits function and how perception works. As Marr put it in a somewhat barbed justification of his new method: 'Trying to understand perception by studying only neurons is like trying to understand bird flight by studying only feathers: it just cannot be done. In order to study bird flight, we have to understand aerodynamics: only then do the structure of the feathers and the different shapes of bird wings make sense.'

Matthew Cobb, The Idea of the Brain

#Matthew Cobb #The Idea of the Brain

1 note · View note

lodane · 5 years ago

Text

[mod list] Star Wars: Knights of the Old Republic - The Sith Lords

I assemled this list for myself and friends/family, but I’m also sharing it with the world in case anyone mods like me and finds this useful. Here’s an excerpt from the convo that sparked this list’s formation as it is...

i have a pretty basic approach to mods for kotor 1&2...

1.) no big cheats (petty stuff -- like slightly beefier stats on armor or something idc, nbd, but i don't install "hack pads," or "skip peragus," things like that)

2.) no weird sex shit. in fact, sometimes i use modesty-mods to combat cringe exploitation.

3.) no heavy-handed mods that disqualify you from using a lot of other mods (so, the massive TSL Restored Content mod is basically off the table, sad to say, sorry... maybe one day i can figure out how to make them all work together) [edit: i have made an exception for TSLRCM, because it’s pretty widely accepted by the community at this point, however this adds SIGNIFICANTLY to the whole install order issue.]

4.) i improve textures on characters, items, effects, skyboxes... but i usually don't mess with super high-res texture replacements of the world/surfaces/deco because 4k textures aren't gonna make KOTOR look THAT much better... however it DOES make your system run like garbage and present more render errors.

5.) gotta be canon (for this game/world) as possible.

- UNOFFICIAL PATCH -

KOTOR 2 Community Patch 1.5.1 https://deadlystream.com/files/file/1280-kotor-2-community-patch/

- BUG FIXES -

Hide Weapons in Animations 🔥 https://www.nexusmods.com/kotor2/mods/53

Darth Malaks Armour https://www.nexusmods.com/kotor2/mods/9

TSL Jedi Malak Mouth Fix 1.1.0 https://deadlystream.com/files/file/1444-tsl-jedi-malak-mouth-fix/

Kotor 1 Texture to Kotor 2 Game Bridge 1.0.0 (note: these textures are the barest necessity, highly suggested even for those that don’t replace textures for quality... but will be overwritten if you load up on HD textures in the latter category) ⚠️ https://deadlystream.com/files/file/1330-kotor-1-texture-to-kotor-2-game-bridge/

JC's Supermodel Fix for K2 1.6 https://deadlystream.com/files/file/1141-jcs-supermodel-fix-for-k2/

Maintenance Officer Realistic Reskin 1.1 https://deadlystream.com/files/file/165-maintenance-officer-realistic-reskin/

Darth Nihilus Animation Fix Update https://www.nexusmods.com/kotor2/mods/1076

Revan Cutscene Forcepower Fix K2 https://www.nexusmods.com/kotor2/mods/71

Kreia's Fall cutscene (in-game) 1.1 https://deadlystream.com/files/file/1228-kreias-fall-cutscene-in-game/

- DIALOGUE CHANGES -

Dahnis Flirt Option for Female PC 1.0 https://deadlystream.com/files/file/1400-dahnis-flirt-option-for-female-pc/

Bao Dur Shield Dialogue Restoration 1.1 https://deadlystream.com/files/file/1206-bao-dur-shield-dialogue-restoration/

Choose Mira or Hanharr 🔥 https://www.moddb.com/mods/the-sith-lords-restored-content-mod-tslrcm/addons/choose-mira-or-hanharr

- PLAYER HEADMORPHS -

Canonical Jedi Exile 1.2 👩 🔥 (note: this is, personally, my fave headmorph.) https://deadlystream.com/files/file/170-canonical-jedi-exile/

PFHA04 Reskin 👩 (dark brown, contour, lips) https://www.nexusmods.com/kotor2/mods/74

TOR Ports: Meetra Surik AKA Jedi Exile Female Player Head for TSL 1.0.1 👩 https://deadlystream.com/files/file/1646-darth333s-ez-swoop-tsl/

Black Haired PMHC04 👨 https://www.nexusmods.com/kotor2/mods/1025

- COMPANION MORPHS & TEXTURES -

Darth Sapien's Presents T3M4 HD 2k 1.00 https://deadlystream.com/files/file/514-darth-sapiens-presents-t3m4-hd-2k/

Darth Sapiens presents: HD 2K Visas Marr 1.0 https://deadlystream.com/files/file/519-darth-sapiens-presents-hd-2k-visas-marr/

4k Atton 1.0 (note: only used clothes, resized to 25%) 🗜️ https://deadlystream.com/files/file/441-4k-atton/

Atton Rand with scruff https://www.nexusmods.com/kotor2/mods/38

Sparkling Mira https://deadlystream.com/files/file/527-sparkling-mira/

TSL Mira Unpoofed 1.0.1 https://deadlystream.com/files/file/1733-tsl-mira-unpoofed/

Mira Shirt Edit https://www.nexusmods.com/kotor2/mods/1075

Player & Party Underwear 2.0 🔥 https://deadlystream.com/files/file/344-player-party-underwear/

- NPC MORPHS & TEXTURES -

Luxa Hair/Body Fix 🔥 (note: INCLUDE outfit change) https://deadlystream.com/files/file/452-luxa-hair-fix/

Darth Sion remake https://www.nexusmods.com/kotor2/mods/984

Darth Sapiens Presents: HD Darth Nihilus 1.00 https://deadlystream.com/files/file/367-darth-sapiens-presents-hd-darth-nihilus/

TSL Better Male Twi'lek Heads 1.3.1 (note: thin-neck version) https://deadlystream.com/files/file/1432-tsl-better-male-twilek-heads/

Malak with Hair (note: this shit’s canon) https://deadlystream.com/files/file/919-malak-with-hair/

- ADDITIONAL ARMORS -

Jedi Journeyman Robes (Luke ROTJ Outfit) 1.0.2 🔥 https://deadlystream.com/files/file/1511-jedi-journeyman-robes-luke-rotj-outfit

TSL Bao-Dur's Charged Armor https://www.nexusmods.com/kotor2/mods/20

TSL Improved Party Outfits 🔥 https://www.nexusmods.com/kotor2/mods/934

Exile Armor (note: greyscaled in photoshop) 🛃 https://www.nexusmods.com/kotor2/mods/4

- MISC MODS -

Party Leveler TSL https://www.nexusmods.com/kotor2/mods/966

Green Level-Up Icon https://www.nexusmods.com/kotor2/mods/26

Remote Tells Influence (note: appears to not currently work with TSLRCM) ⚠️ https://www.gamefront.com/games/knights-of-the-old-republic-ii/file/remote-tells-influence

Invisible Headgear 🔥 https://deadlystream.com/files/file/736-invisible-headgear/

Darth333's EZ Swoop [TSL] 1.0.0 (note: players say it does currently work with TSLRCM) https://deadlystream.com/files/file/1646-darth333s-ez-swoop-tsl/

- HD TEXTURE UPGRADES -

note: many of these textures overlap and will overwrite one another. i haven’t listed them in any particular order... except the first two. i suggest, if you really want almost every texture in the game to be upgraded, to install them first and let everything else overwrite according to your prefs. if you wanna keep it simple, i suggest just skipping this and the next category (loading screens)

...

*KOTOR 2 UNLIMITED WORLD TEXTURE MOD (note: resized to 25%) 🗜️ https://www.nexusmods.com/kotor2/mods/1062

*Complete Character Overhaul - Ultimate HD Pack (note: resized to 25%) 🗜️ https://www.nexusmods.com/kotor2/mods/1060

High Quality Blasters https://www.nexusmods.com/kotor2/mods/919

HD Lightsabers https://www.nexusmods.com/kotor2/mods/1013

The IceEclipse Power Textures https://www.nexusmods.com/kotor2/mods/960

Fire and Ice HD https://www.nexusmods.com/kotor2/mods/25

HD Foot Locker https://www.nexusmods.com/kotor2/mods/990

TSL Backdrop Improvements https://www.nexusmods.com/kotor2/mods/922

Improved Peragus Asteroid Fields 1.2 https://deadlystream.com/files/file/321-improved-peragus-asteroid-fields/

Telos Polar Academy Hangar Skybox 1.0.0 https://deadlystream.com/files/file/1389-telos-polar-academy-hangar-skybox/

High Quality Ravager Backdrop https://www.nexusmods.com/kotor2/mods/41

High Quality Stars and Nebulas https://www.nexusmods.com/kotor2/mods/31

Sith Soldier armor retexture https://www.nexusmods.com/kotor2/mods/1037

High Poly Grenades K2 https://www.nexusmods.com/kotor2/mods/1001

HD Muzzle Flash https://www.nexusmods.com/kotor2/mods/991

KotOR 2 Remastered (AI Upscaled) Cutscenes (note: 720p) https://www.nexusmods.com/kotor2/mods/1066

K2 - Upscaled maps 🔥 https://www.nexusmods.com/kotor2/mods/1086

Atris Holocrons https://www.nexusmods.com/kotor2/mods/949

TSL HD Cockpit Skyboxes 3.1 (High Res TPC) https://deadlystream.com/files/file/931-tsl-hd-cockpit-skyboxes/

More Vibrant Skies 1.0 https://deadlystream.com/files/file/156-more-vibrant-skies/

Vurt's Exterior Graphics Overhaul https://www.nexusmods.com/kotor2/mods/1044

Ebon Hawk HD - 4X Upscaled Texture (note: resized to 25%) 🗜️ https://www.nexusmods.com/kotor2/mods/1059

Ebon Hawk Model Fixes 2.0 https://deadlystream.com/files/file/1033-ebon-hawk-model-fixes/

Peragus Administration Level Room Model Fixes 1.0.0 https://deadlystream.com/files/file/1275-peragus-administration-level-room-model-fixes/

Peragus Large Monitor Adjustment 1.0 https://deadlystream.com/files/file/317-peragus-large-monitor-adjustment/

[TSL] Animated Computer Panel 1.1.0 https://deadlystream.com/files/file/1385-tsl-animated-computer-panel/

Citadel Station Signage 1.1 https://deadlystream.com/files/file/308-citadel-station-signage/

Replacement Peragus II Artwork by Trench 1.0 https://deadlystream.com/files/file/361-replacement-peragus-ii-artwork-by-trench/

Replacement Texture for Lightning on Malachor V 1.2 https://deadlystream.com/files/file/298-replacement-texture-for-lightning-on-malachor-v/

Peragus Medical Monitors and Computer Panel 1.0.0 https://deadlystream.com/files/file/1375-peragus-medical-monitors-and-computer-panel/

TSL Animated Galaxy Map 4.1 https://deadlystream.com/files/file/219-tsl-animated-galaxy-map/

- HD LOADING SCREENS -

Replacement Loading Screens for KotOR2: Original Pack (with or without TSLRCM) - Part 1 1.1 https://deadlystream.com/files/file/1255-replacement-loading-screens-for-kotor2-original-pack-with-or-without-tslrcm-part-1/

Replacement Loading Screens for KotOR2: Original Pack (with or without TSLRCM) - Part 2 1.1 https://deadlystream.com/files/file/1256-replacement-loading-screens-for-kotor2-original-pack-with-or-without-tslrcm-part-2/

Replacement Loading Screens for KotOR2: Original Pack (with or without TSLRCM) - Part 3 1.0 https://deadlystream.com/files/file/1257-replacement-loading-screens-for-kotor2-original-pack-with-or-without-tslrcm-part-3/

Replacement Loading Screens for KotOR2: Add-On Pack (with or without TSLRCM) 1.3 https://deadlystream.com/files/file/1253-replacement-loading-screens-for-kotor2-add-on-pack-with-or-without-tslrcm/

Nar Shaddaa Loadscreens (using Sharen Thrawn skybox and with horizontal overlay) 1.0 https://deadlystream.com/files/file/1238-nar-shaddaa-loadscreens-using-sharen-thrawn-skybox-and-with-horizontal-overlay/

- UNOFFICIAL RESTORATIONS -

The Sith Lords Restored Content Mod (TSLRCM) https://www.moddb.com/mods/the-sith-lords-restored-content-mod-tslrcm

Extended Enclave (TSLRCM add-on) https://www.moddb.com/mods/the-sith-lords-restored-content-mod-tslrcm/addons/extended-enclave-tslrcm-add-on

NPC Overhaul Mod (TSLRCM add-on) (note: this is an inane mess to install, tedious AF, and idk if i can say it’s worth it... but i did it and it didn’t break anything) https://www.moddb.com/mods/the-sith-lords-restored-content-mod-tslrcm/addons/npc-overhaul-mod

Party Swap (TSLRCM add-on) 🔥 https://www.moddb.com/mods/partyswap/news/partyswap-133

Coruscant Scene No Overlay (TSLRCM add-on) https://www.moddb.com/mods/the-sith-lords-restored-content-mod-tslrcm/addons/coruscant-scene-no-overlay

#kotor 2 #mod lists

5 notes · View notes

shijinpunnaykaljoseph · 6 years ago

Text

#Blog Post 9. ARTIFICIAL INTELLIGENCE AND MUSIC

Music is undoubtedly one of the most intriguing human intelligence activities. For composers, artists, authors, engineers and scientists, computers are a fantastic tool. All of these give music a different perspective, but all of them can benefit from computers ' precision and automation. Computers have more to offer than just a decent calculator, printer, audio recorder, digital effects box, or whatever the computer's immediate useful purpose. The computer's real significance lies in the new paradigms of art and scientific thinking that computation creates. The technology of using Artificial Intelligence for creating music is latest innovations in the field of technology. Discussion days are over when artificial intelligence (AI) will impact the music industry. In many ways, artificial intelligence is already being used. Today, in music composition, production, theory and digital sound processing, you can find AI applications. In addition, AI helps musicians test new ideas, find the best emotional context, incorporate music into modern media, and just have fun. In huge data sets, AI automates processes, finds trends and observations, and helps create efficiencies.

Into the history

First phase

Work at this time focused primarily on algorithmic composition which aimed at a new composition that is aesthetically satisfying:

In 1957, Lejaren Hiller and Leonard Isaacson of Urbana's University of Illinois programmed “Illiac Suite for String Quartet”, the first work entirely written by artificial intelligence.

In 1960, the first paper on algorithmic music composition was written by Russian researcher R.Kh. Zaripov using the "Ural-1" computer.

Breakthroughs

The more significant level of music intelligence emerged in the generative modelling of music as research began to focus on understanding music.

In the 1970s, interest in algorithmic music also reached the pop scene's well-known artists. David Bowie, an unquestionable iconic figure in the music industry, was the one who first started thinking in this direction. He created Verbasizer, a lyric-writing Mac app, with Ty Roberts.

In 1975, it was N. Rowe from the MIT Experimental Music Studio has created a smart music perception system that allows a musician to play freely on an acoustic keyboard while the computer infers a metre, its tempo and note durations.

In 1980, David Cope founded EMI (Experiments in Musical Intelligence) from the University of California, Santa Cruz. The system was based on generative models for analysing and creating new pieces based on existing music.

Current research

AI and music research continues on composition of music, intelligent analysis of sound, cognitive science and music, etc.

Initiatives like Google Magenta, Sony Flow Machines, IBM Watson Beat, would like to find out whether AI can compose music that is convincing.

Pop music composed by AI was first published by the Sony CSL Research Laboratory in 2016 and the results were impressive. Daddy's Car's is a cheerful, bright tune that resembles The Beatles. The AI system offered to write the song melodies and lyrics based on the original parts of The Beatles, but the further arrangement was rendered by live musicians.

Taryn Southern became a sensation after taking part in the American Idol TV show. The next logical step was for a new album to be released. Southern decided to take a non-standard approach — to use AI to write an album. She opted for a startup called Amper as a tool. This programme is capable of producing sets of melodies according to a given genre and mood with the help of internal algorithms. The outcome was the “Break Free” song.

CJ Carr and Zack Zukowski are engaged in a very unusual thing founded by Dadabots Boston programmers— they teach artificial intelligence how to write "heavy" music. The developers presented the black metal album "Coditany of Timeness" in 2017 and presented the results of the algorithm that composes music in the death metal style to the public. According to them, the algorithm produces a pretty decent music for this genre, which needs no corrections, so they decided to give him the willingness to write live tracks on YouTube.

Interesting developments in the Musical Intelligence field

Research now focuses on using artificial intelligence in compiling musical composition, performance, and digital sound processing, as well as selling and consuming music. For teaching and creating music, most AI-based systems and applications are used. Some of them are here:

AlgoTunes is a software company that develops music-generating applications. On the site, with one keystroke, anyone can create a random piece of music with a given style and mood— but the choice of settings is very limited. Music is created in a few seconds by a web application and can be downloaded as WAV or MIDI files.

Founded in 2015, MXX (Mashtraxx Ltd) is the world's first artificial intelligence mechanism to instantly convert music to video using a stereo file. MXX enables you to adapt music to specific user content, such as sports and playing, computer games plots, and so on. Audition pro, the first MXX app, allows anyone to edit music for video: load an existing song and automatically adjust the increase in sound frequency, amplification and pause according to the video's dynamics. MXX also provides services to leading commercial libraries, music services, game developers, and production studios that include music tailored to modern media.

Orb Composer — a program designed by Hexachords to help compile orchestral compositions at genre selection stages, instrument selection, track composition.

OrchExtra may help collect a complete Broadway score from a small high school or city theatre ensemble. OrchExtra plays the role of missing devices, recording variations in tempo and musical language.

Is AI a threat to musicians?

AI's capabilities create tension among the groups of musicians and producers who see it as a challenge to their jobs first. AI-music start-ups ' argument is that AI is a creative tool that frees musicians to create more and better art. While AI can outperform humans to help you sleep in areas like video backup music or soundtracks, it is not capable of creating great original music without human intervention.

Conclusion

AI will eventually change the music industry as computers and AI become more powerful and accessible. The choice is how we work with AI, though. The reason that robots can take over our jobs is that we are robotic about the jobs. However the creation and production of music is a creative process. After all, the possible mechanism of typically developing legendary songs happens on its own, from deep within the heart of the poet, his passionate feelings and unique experience of life. Giving more control over AI applications to musicians is important rather than letting AI take over. After all, we love musicians because of their humanity and personality — not just their music itself.

References

1. Deahl, D. (2018). How AI-generated music is changing the way hits are made. [Online] The Verge. Available at: https://www.theverge.com/2018/8/31/17777008/artificial-intelligence-taryn-southern-amper-music [Accessed 9 Nov. 2019].

2. Drake, J. (2018). AI & Music. [Online] Soundonsound.com. Available at: https://www.soundonsound.com/music-business/ai-music [Accessed 21 Nov. 2019].

3. En.wikipedia.org. (2019). Music and artificial intelligence. [Online] Available at: https://en.wikipedia.org/wiki/Music_and_artificial_intelligence [Accessed 21 Nov. 2019].

4. Kharkovyna, O. (n.d.). Artificial Intelligence and Music: What to Expect? [Online] Medium. Available at: https://towardsdatascience.com/artificial-intelligence-and-music-what-to-expect-f5125cfc934f [Accessed 21 Nov. 2019].

5. Li, C. (2019). A Retrospective of AI + Music. [Online] Medium. Available at: https://blog.prototypr.io/a-retrospective-of-ai-music-95bfa9b38531 [Accessed 21 Nov. 2019].

6. Love, T. (2019). Do androids dream of electric beats? How AI is changing music for good. [Online] the Guardian. Available at: https://www.theguardian.com/music/2018/oct/22/ai-artificial-intelligence-composing [Accessed 21 Nov. 2019].

7. Marr, B. (2019). The Amazing Ways Artificial Intelligence Is Transforming The Music Industry. [Online] Forbes.com. Available at: https://www.forbes.com/sites/bernardmarr/2019/07/05/the-amazing-ways-artificial-intelligence-is-transforming-the-music-industry/#23c4c1a65072 [Accessed 21 Nov. 2019].

1 note · View note

whiskasgirl · 6 years ago

Text

KOTOR 2 Mod list

For personal reference. Please feel free to add to this if you have further recommendations!! :) I love this game so much istg

Widescreen Fix

Reshade

Gameplay

TSLRCM M4-78EP Extended Enclave NPC Overhaul Mod Sion's arrival at Peragus - High Resolution Movies Nihilus Zombies Death CORUSCANT - JEDI TEMPLE Coruscant - Jedi Temple Compatibility For M4-78EP Peragus Sith Troops To Sith Assassins My Dantooine Statue Power Cost Correction Create all Mira's Rockets Full Jedi Council TSL Loot & Immersion Upgrade v.3b (caused crashing) TSL Improved Party Outfits The Exile's Robe Holowan Duplisaber Beta .215b Casino PAZAAK animated 1.2 SWKOTOR2 Secret Tomb Bugfix

Textures&models

KOTOR 2 Starter Pack Kotor 1 Texture to Kotor 2 Game Bridge 1.0.0 Textures Improvement Project 0.4.1 Replacement Loading Screens for KotOR2: Original Pack (Part 1) Replacement Loading Screens for KotOR2: Original Pack (Part 2) Classic Jedi Project KOTOR II Original Trilogy Lightsaber Blades (Animated) Classic Jedi Project KOTOR II Original Trilogy Lightsaber Sounds High Quality Blasters The IceEclipse Power Textures My Realistic Improved Effect Retexture Mod 1.0 Animated Galaxy Map High Quality Stars and Nebulas Movie-style Jedi Master Robes Sith Soldier Armor Retexture Combat Suit Revisited 4.2 Fixed Hologram Models and Admiralty Redux for TSLRCM 1.61 [TSL] Animated Computer Panel 1.1.0 TSL Backdrop Improvements HD Foot Locker TSL Twi'lek Male 3D Ears 1.2.1 Improved Jedi Sacks TSL Harbinger and Hammerhead Visual Appearance Mod 1.0

Places

Peragus OTE & A Darker Peragus + A Darker Peragus REDUX 1.0 Improved Peragus Asteroid Fields Peragus Large Monitor Adjustment 1.0 Peragus Medical Monitors and Computer Panel 1.0.0 TSL ORIGINS - Telos Overhaul 1.1.1 & Telos OTE / K2TSLR - The Sith Lords Remastered WIP Alpha 1-1 TSL ORIGINS: Harbinger Overhaul 1.0 New Texture of Holocrons in the Telos Secret Academy 1.2 Quansword's HD Ebon Hawk Retexture Ebon Hawk Past KOTOR2 More Subtle Animated Ebon Hawk Monitors (still not including Galaxy Map) 1.2 Dantooine and Korriban High Resolution Improved Dantooine Skybox 1.0.1 DXUN 2013 K2 Exterior Textures, Part 1 Version 0.1 Vurt’s Exterior Graphics Overhaul Narshadaa O.T.E. Realistic Nar Shaddaa Skybox HD Skybox for M4-78 Malachor Exterior Reskin By Quanon V1.3 Replacement Texture for Lightning on Malachor V 1.2 Side Opening Doors on Malachor Realistic Skybox for deathdisco's Coruscant 1.0.0 More Vibrant Skies 1.0

Characters

Party Member Reskin Mod (Kreia, Bao-Dur, Desciple, Hanharr) Atton Reskin 4K Atton (outfit) Sparkling Mira Darth Sapiens presents: HD 2K Visas Marr 1.0 Darth Sapien's Presents T3M4 HD 2k 1.00 HK-50 & HK-51 Reskin Movie Mandalorians TFU-style Darth Sion model Darth Sapiens Presents: HD Darth Nihilus 1.00 Visually Repair HK-47 1.0.0 Taibhrigh's Female Player Head Replacement Skin (PFHC07) PFHA04 Reskin Luxa Hair Fix Maintenance Officer Realistic Reskin 1.1

#kotor2 #kotor mods #for future reference #long ass post alert

9 notes · View notes

thepissprophet · 8 months ago

Text

The way direct realism was described to me made it seem very hippy

JJ Gibson's ideas being compared and contrasted to the computational model guys (Marr and Bierderman) really did it for me.

Like computational models give me the vibe of the ultimate pursuit of pure science. Break everything down into this clean, sterile and categorisation of things.

Where direct realism is like, nature is chaotic. Our psychology is chaotic. You wanna study leaf litter ecology? You can't study each animal in isolation and know what would happen when all the biotic and abiotic elements are placed together! This can't be expressed with nice mathematical models

(Nature can oft be studied using mathematical models yes. But they get so big and complicated just to try and account for all the different factors lol)

#direct idealism #J J Gibson #Marr's computational model #cogntive psychology #poorly thought out psychology rant

0 notes

myfriendpokey · 7 years ago

Text

promissory notes

Complete satisfaction OR your money back!!! - is something I don't think anyone in videogames has ever willingly said,so maybe it doesn't make sense to talk of anything as stable as a "guarantee". Maybe it's more like a system of overlapping promises, designed to contain the idea that a videogame exists in at least some kind of provisional relationship to human happiness, even if the rate at which the two could be converted is never quite nailed down. We have the promise the game will work on a given system or set of minimum requirements, the hazier assurance it might at least resemble the screenshots on the box, the genre assurance that it formally and hence experientially resembles some other game you like. The assurance, in press leading up to release, of passion and artistic intent rattling around in there somewhere as well, the assurance that the game will have x y and z new features and scope. Press and external reviewers can so to speak cosign a guarantee or write their own more ambivalent one on the basis of their reputation. Storefronts as well can tacitly endorse some promise - that this thing exists, functions, falls into the category of "entertainment" - when they put it on their shelves, virtual or otherwise. There's the promise of personal reputation, that the people involved wouldn't want to associate themselves with a bogus product, and the promise of monetary interest - this game obviously had a fair bit of money put into it, they're expecting to make that back, therefore we can expect some moderate fidelity to customer expectation and the sort of general polished feel that comes with being able to hire lots of people to create bark textures.

Most of these institutions aren't specific to videogames, but I do think they have a greater prominence there, owing both to the higher amount of fussy technical variation in the format (it's hard to imagine a book, say, refusing to boot or secretly installing a bitcoin miner in your head) and also to its historical novelty. The idea that something called a "videogame" exists, is an entertainment format, is linked to some kind of prospective emotional value - all these have to be rhetorically insisted upon, particularly as the format moved from spaces with immediately visible analogues (pinball tables, mechanical amusements) to a more diffuse place alongside the family television or home computer. They had to insinuate, and to an extent still have to insinuate, the exact role they played in everyday life. And the shift from being a sort of weird, garish, once-off toy into an ongoing home-improvement project, with new games and consoles to choose between and new add-ins to improve your machine, had to be accompanied by the emergence of institutions that could offer some reassurance this ongoing investment wouldn't be a waste.

So you can maybe glibly think of videogames as a form of currency, built on the premise that they can be "exchanged" at any time for some measure of enjoyment, where this exchange rate is underwritten and co-signed by various institutions. And as having something of the abstraction of currency, as well. If one videogame is a moment of enjoyment then 6000 videogames are in principle 6000 moments of enjoyment - never mind that you may never have a chance to play all those discounted games in your Steam library within one lifetime. Think of it as saving them up for a rainy day. And I suspect that as this relationship between possession and affect grows more abstracted and tenuous, institutions take on a correspondingly more important role in confirming that the central exchange relationship still holds true. A bit like debt rating agencies - it's not so much about actually untangling the complicated sale of good, bad and nonexistent debt packages from one financial institution to another, it's more the promise that at some point this untangling COULD occur, that all this imaginary money still bears some kind of distant relationship to actual human needs.

I wonder if the paranoid style in videogames culture stems partly from this sense of underlying contingency. It's not that games are just experiences, which can't be taken away - they're more like deposits on hypothetical experience, and those deposits can indeed depreciate in value if not turn out to be worthless from the start. Bad reviews, spoilers, the general reputation of a game can all cause it to drop in expected value. The fuss that happens every time a new GTA game gets below 9.5 on IGN or wherever is not so much that the game might really have problems so much that having those problems flagged from the start can marr the sense of occasion, the I-was-there-ness and anticipated retrospective value that's part of the package being sold. And of course the consistent anxiety around corrupt reviewers, incorrect press releases, "fake games", all those other things that could adulterate the currency...

And maybe we could consider the current anxiety around "asset flips" on Steam in the same light. After all, who's really playing these things - besides Youtubers doing so ironically? They're easy to spot and easily refunded and even if some kid really does buy "Cuphat" or "Battleglounds" by accident, well, the worst that could happen is that they develop the same misplaced affection for exploitative consumer garbage as everybody else who grew up playing videogames. And indeed the fact that nobody really buys them is part of the critique - what's unsettling is the fact that they seem more connected to the shadow economies of cheap bundles and trading-card-store manipulation (which is so easy and widespread that PC Gamer could publish a how-to guide with no apparent pushback from anyone). You can easily unpick the specific arguments about what constitutes an asset flip versus a game that just uses premade assets, or how to tell a "scam" from just a regular bad videogame - demonstrative sincerity?? Producing cynical knockoffs with premade asset packs is not necessarily the act of poorly-funded fly-by-nights, as witness the recent news about Voodoo recieving $200 million from Goldman Sachs. But of course they're the chief source of anxiety around the issue, and the ways in which that anxiety manifests is often weirdly racialised - the automatic bad faith extended to the Global South, the fear of nameless hordes overrunning our valuable, exclusive institutions, even a sort of weird variant on the “welfare mother” imagery - the asset flipper with 100 interchangeable games, driving a cadillac... Leaching off the accumulated value stockpiled by the Steam brand, devaluing our libraries and the institutions that have been telling us they're worth something...

I don't really have a lot of sympathy with the asset-flip discourse, both because exactly the same anxiety has been rolled out in the past to Unity, walking simulators, visual novels, Game Maker, Twine, and basically anything else that lowers the barriers to entry around making videogames; and also because I love many games I think those anxieties would try to exclude ("The Zoo Race", GoreBagg games, the Johnny series, even Limbo Of The Lost is as close as Oblivion ever came to being creatively exciting), and I think the calls for hard work and sincerity and so forth function as just so much evasive kitsch. We already HAVE a ton of games like that; and that's maybe the real problem. Why is there so much anxiety about discovering good games when, say, people are complaining about having to choose between the two different, polished, labour-of-love, years-in -the-making narrative platform games being released the same week? Doesn't this just mean the "enjoyment standard" of the videogame promissory note is just by now so abstract and intangible that it's basically just an empty convention, useful for nothing but perpetuating itself - perpetuating the idea of an unadulterated good-game-ness, stretching aimlessly into space like a 1950s radio broadcast. It's a convention which is basically exclusive, which works by trying to put a cordon around the vast swathes of human culture it thinks it's safe to ignore.

Which is maybe fine - nobody can pay attention to anything, and some "rating institutions" are presumably less pernicious than others (the advice of a friend? a critic you enjoy? your own intuition?). There are obviously a lot of critiques that can be levelled at the existing one for videogames, including in particular the assumption that anything that cost a lot of money is worth at least checking out. But there's also something more generally sad about this kind of enforced, perpetual scarcity in a time of abundance, about a model that just pines for less shit so that it can start to feel relevant again, about one that can think of nothing to do with the sheer volume of things being made and rabbit holes being burrowed than wish they didn't exist and try to shut them out entirely.

More people being able to make things is good, and hard to get to; it can also be unnerving and disorienting and also push against some of the happier ideas we might have had about the democratization of art-production (for example, that this wouldn't co-exist with monopolies of arbitrary unaccountable control of the kind exercised by Youtube, Steam, the App store, Google, etc...), it can be a space to view some of the weirder machinations of capital as they leave traces through the culture (money-laundering $9000 books on Amazon and viral Pregnant Spiderman youtube vids). I don't think continuing to defend the value of the medium will help think about these, or become anything but more and more paranoid and quixotic over time.

#lou #vgames #money

34 notes · View notes

compneuropapers · 2 years ago

Text

Interesting Papers for Week 10, 2023

A computational model to explore how temporal stimulation patterns affect synapse plasticity. Amano, R., Nakao, M., Matsumiya, K., & Miwakeichi, F. (2022). PLOS ONE, 17(9), e0275059.

Distinct population and single-neuron selectivity for executive and episodic processing in human dorsal posterior cingulate. Aponik-Gremillion, L., Chen, Y. Y., Bartoli, E., Koslov, S. R., Rey, H. G., Weiner, K. S., … Foster, B. L. (2022). eLife, 11, e80722.

Hippocampal astrocytes encode reward location. Doron, A., Rubin, A., Benmelech-Chovav, A., Benaim, N., Carmi, T., Refaeli, R., … Goshen, I. (2022). Nature, 609(7928), 772–778.

Classical center-surround receptive fields facilitate novel object detection in retinal bipolar cells. Gaynes, J. A., Budoff, S. A., Grybko, M. J., Hunt, J. B., & Poleg-Polsky, A. (2022). Nature Communications, 13, 5575.

Inferring stimulation induced short-term synaptic plasticity dynamics using novel dual optimization algorithm. Ghadimi, A., Steiner, L. A., Popovic, M. R., Milosevic, L., & Lankarany, M. (2022). PLOS ONE, 17(9), e0273699.

Context-dependent selectivity to natural images in the retina. Goldin, M. A., Lefebvre, B., Virgili, S., Pham Van Cang, M. K., Ecker, A., Mora, T., … Marre, O. (2022). Nature Communications, 13, 5556.

Dendritic axon origin enables information gating by perisomatic inhibition in pyramidal neurons. Hodapp, A., Kaiser, M. E., Thome, C., Ding, L., Rozov, A., Klumpp, M., … Both, M. (2022). Science, 377(6613), 1448–1452.

Long‐term modulation of the axonal refractory period. Jankowska, E., Kaczmarek, D., & Hammar, I. (2022). European Journal of Neuroscience, 56(7), 4983–4999.

Flexible control as surrogate reward or dynamic reward maximization. Liljeholm, M. (2022). Cognition, 229, 105262.

Mixed synapses reconcile violations of the size principle in zebrafish spinal cord. Menelaou, E., Kishore, S., & McLean, D. L. (2022). eLife, 11, e64063.

Adolescents sample more information prior to decisions than adults when effort costs increase. Niebaum, J. C., Kramer, A.-W., Huizenga, H. M., & van den Bos, W. (2022). Developmental Psychology, 58(10), 1974–1985.

Neural signatures of evidence accumulation in temporal decisions. Ofir, N., & Landau, A. N. (2022). Current Biology, 32(18), 4093-4100.e6.

Offset responses in the auditory cortex show unique history dependence. Olsen, T., & Hasenstaub, A. (2022). Journal of Neuroscience, 42(39), 7370–7385.

Look-up and look-down neurons in the mouse visual thalamus during freely moving exploration. Orlowska-Feuer, P., Ebrahimi, A. S., Zippo, A. G., Petersen, R. S., Lucas, R. J., & Storchi, R. (2022). Current Biology, 32(18), 3987-3999.e4.

Interaction of bottom-up and top-down neural mechanisms in spatial multi-talker speech perception. Patel, P., van der Heijden, K., Bickel, S., Herrero, J. L., Mehta, A. D., & Mesgarani, N. (2022). Current Biology, 32(18), 3971-3986.e4.

Spatial distances affect temporal prediction and interception. Schroeger, A., Grießbach, E., Raab, M., & Cañal-Bruland, R. (2022). Scientific Reports, 12, 15786.

Never run a changing system: Action-effect contingency shapes prospective agency. Schwarz, K. A., Klaffehn, A. L., Hauke-Forman, N., Muth, F. V., & Pfister, R. (2022). Cognition, 229, 105250.

Contraction bias in temporal estimation. Tal-Perry, N., & Yuval-Greenberg, S. (2022). Cognition, 229, 105234.

Response Time Distributions and the Accumulation of Visual Evidence in Freely Moving Mice. Treviño, M., Medina-Coss y León, R., & Lezama, E. (2022). Neuroscience, 501, 25–41.

Active neural coordination of motor behaviors with internal states. Zhang, Y. S., Takahashi, D. Y., El Hady, A., Liao, D. A., & Ghazanfar, A. A. (2022). Proceedings of the National Academy of Sciences, 119(39), e2201194119.

#science #Neuroscience #computational neuroscience #Brain science #research #cognition #cognitive science #psychophysics #neurons #neural networks #neurobiology #neural computation #scientific publications

5 notes · View notes

jcmarchi · 2 years ago

Text

Using language to give robots a better grasp of an open-ended world

New Post has been published on https://thedigitalinsider.com/using-language-to-give-robots-a-better-grasp-of-an-open-ended-world/

Using language to give robots a better grasp of an open-ended world

Imagine you’re visiting a friend abroad, and you look inside their fridge to see what would make for a great breakfast. Many of the items initially appear foreign to you, with each one encased in unfamiliar packaging and containers. Despite these visual distinctions, you begin to understand what each one is used for and pick them up as needed.

Inspired by humans’ ability to handle unfamiliar objects, a group from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) designed Feature Fields for Robotic Manipulation (F3RM), a system that blends 2D images with foundation model features into 3D scenes to help robots identify and grasp nearby items. F3RM can interpret open-ended language prompts from humans, making the method helpful in real-world environments that contain thousands of objects, like warehouses and households.

F3RM offers robots the ability to interpret open-ended text prompts using natural language, helping the machines manipulate objects. As a result, the machines can understand less-specific requests from humans and still complete the desired task. For example, if a user asks the robot to “pick up a tall mug,” the robot can locate and grab the item that best fits that description.

“Making robots that can actually generalize in the real world is incredibly hard,” says Ge Yang, postdoc at the National Science Foundation AI Institute for Artificial Intelligence and Fundamental Interactions and MIT CSAIL. “We really want to figure out how to do that, so with this project, we try to push for an aggressive level of generalization, from just three or four objects to anything we find in MIT’s Stata Center. We wanted to learn how to make robots as flexible as ourselves, since we can grasp and place objects even though we’ve never seen them before.”

Learning “what’s where by looking”

For example, the fulfillment centers of major online retailers can contain millions of items, many of which a robot will have never encountered before. To operate at such a scale, robots need to understand the geometry and semantics of different items, with some being in tight spaces. With F3RM’s advanced spatial and semantic perception abilities, a robot could become more effective at locating an object, placing it in a bin, and then sending it along for packaging. Ultimately, this would help factory workers ship customers’ orders more efficiently.

“One thing that often surprises people with F3RM is that the same system also works on a room and building scale, and can be used to build simulation environments for robot learning and large maps,” says Yang. “But before we scale up this work further, we want to first make this system work really fast. This way, we can use this type of representation for more dynamic robotic control tasks, hopefully in real-time, so that robots that handle more dynamic tasks can use it for perception.”

“Visual perception was defined by David Marr as the problem of knowing ‘what is where by looking,’” says senior author Phillip Isola, MIT associate professor of electrical engineering and computer science and CSAIL principal investigator. “Recent foundation models have gotten really good at knowing what they are looking at; they can recognize thousands of object categories and provide detailed text descriptions of images. At the same time, radiance fields have gotten really good at representing where stuff is in a scene. The combination of these two approaches can create a representation of what is where in 3D, and what our work shows is that this combination is especially useful for robotic tasks, which require manipulating objects in 3D.”

Creating a “digital twin”

Keeping things open-ended

After receiving a few demonstrations, the robot applies what it knows about geometry and semantics to grasp objects it has never encountered before. Once a user submits a text query, the robot searches through the space of possible grasps to identify those most likely to succeed in picking up the object requested by the user. Each potential option is scored based on its relevance to the prompt, similarity to the demonstrations the robot has been trained on, and if it causes any collisions. The highest-scored grasp is then chosen and executed.

F3RM also enables users to specify which object they want the robot to handle at different levels of linguistic detail. For example, if there is a metal mug and a glass mug, the user can ask the robot for the “glass mug.” If the bot sees two glass mugs and one of them is filled with coffee and the other with juice, the user can ask for the “glass mug with coffee.” The foundation model features embedded within the feature field enable this level of open-ended understanding.

“If I showed a person how to pick up a mug by the lip, they could easily transfer that knowledge to pick up objects with similar geometries such as bowls, measuring beakers, or even rolls of tape. For robots, achieving this level of adaptability has been quite challenging,” says MIT PhD student, CSAIL affiliate, and co-lead author William Shen. “F3RM combines geometric understanding with semantics from foundation models trained on internet-scale data to enable this level of aggressive generalization from just a small number of demonstrations.”

Shen and Yang wrote the paper under the supervision of Isola, with MIT professor and CSAIL principal investigator Leslie Pack Kaelbling and undergraduate students Alan Yu and Jansen Wong as co-authors. The team was supported, in part, by Amazon.com Services, the National Science Foundation, the Air Force Office of Scientific Research, the Office of Naval Research’s Multidisciplinary University Initiative, the Army Research Office, the MIT-IBM Watson Lab, and the MIT Quest for Intelligence. Their work will be presented at the 2023 Conference on Robot Learning.

0 notes

tak4hir0 · 5 years ago

Link

You don't have to spend a fortune and study for years to start working with big data, analytics, and artificial intelligence. Demand for "armchair data scientists" – those without formal qualifications in the subject but with the skills and knowledge to analyze data in their everyday work, is predicted to outstrip demand for traditionally qualified data scientists in the coming years. The 9 Best Free Online Data Science Courses In 2020 Adobe Stock This means that practically anyone can upgrade their employability and career prospects by learning the fundamental theory and practical skills needed for data science. And luckily, there's a ton of resources online to help you do just that. Here’s my overview of some of the best. Some of these might require payment at the end of the course if you want official certification or accreditation of completing the course, but the learning material is freely available to anyone who wants to level up their data knowledge and skills. Data Science Crash Course, John Hopkins University (Coursera) Designed to give a "fluff-free" overview of what data science is, how it works, and what it can be used to do. This course offers an introduction to the technical side of data science but is particularly aimed at understanding the "big picture" for those who need to manage data scientists or data science work. It's a relatively short course consisting of just one module that can be completed in under a week and serves as a great introduction for those who want to learn the terminology and understand how to build a data science strategy, without necessarily needing detailed instructions on using the technical tools involved. Introduction to Data Science (Revised) - Alison A completely free course that breaks down the core topics of the data science process and an introduction to machine learning into three modules, each designed to take around three hours to complete, and concluding with an assessment. Once you've worked through that, you can choose from several other similarly bite-sized tutorials covering data programming languages, visualization tools, and techniques such as building clustering and regression models. Data Science and Machine Learning Essentials – Microsoft (EdX) This course, aimed at those wanting to improve their career prospects with a mix of practical and theoretical knowledge, walks you through core concepts and terminology, statistical techniques such as regression, clustering, and classification, and the practical steps needed to build and evaluate models. As it is a Microsoft course, its cloud-based components focus on the company's Azure framework, but the concepts that are taught are equally applicable in organizations that are tied to competing cloud frameworks such as AWS. It assumes a basic understanding of R or Python, the two most frequently used programming languages in data science, so it may be useful to look at one of the courses covering those that are mentioned below, first. Learn Data Science - Dataquest Although primarily a paid-for platform offering proprietary content, Dataquest offers a number of free introductory modules to anyone who signs up, covering essential topics such as working with data, visualizing data, data mining and constructing algorithms in Python and R. If you want the full, ad-free experience and certification there are monthly subscription options, but there's more than enough information to get started free of charge. Data Science - Harvard All of the class materials and lectures for Harvard's data science course are made freely available online, so they can be studied at your own pace. You may not end up with a degree from one of the world's most prestigious universities, but the course is detailed and technical enough to make an expert of you by the end. The course is part of a data science degree and constructed for students who have prior knowledge of, or are also studying, core fields such as programming, maths, and statistics. However, there are enough free resources out there on those subjects to make this a viable option for those outside of academia, if you are dedicated enough. Introduction to Data Science in Python – University of Michigan (Coursera) Those wanting to get their hands dirty with some actual coding will soon find out that Python is one of the most commonly used programming languages in the field, and for good reason. It’s relatively simple to learn the basics and can be combined with a number of free, open-source libraries to perform hugely powerful data science operations. This course serves as a first step along the road, introducing Python functions that are used to prepare and manipulate big datasets as well as the proven techniques for extracting insights from data. It is intended to be completed by spending between three and six hours per week studying or working on exercises, over four weeks. Learn Data Science with R – Ram Reddy (Coursera) This course led by an established expert in R and data analytics is the first in an in-depth, ten-part tutorial on expert R programming, but also stands on its own as an introduction to the language and a primer on the basics as they relate to data science. Like Python, R is a totally free and open-source language and environment that has become an accepted standard among data scientists due to its power and flexibility. This course consists of 10 lectures delivered across eight hours of video, and is completely free to follow. Introduction to Data Science Using Python - Rakesh Gopalakrishnan (Udemy) This is one of the most highly-rated of Udemy’s introductory courses on the subjects of data science and coding in Python. It does not require any previous knowledge or experience as it starts right from the basics. However, unlike some other very entry-level courses, it does progress to some actual practical instruction in Python and, particularly usefully, its Sci-Kit Learn framework, a very popular tool for academic and enterprise-level data exploration and mining. I Heart Stats: Learning to Love Statistics – University of Notre Dame (EdX) Along with maths and computer science, statistics is one of the fundamental academic disciplines invoked by those working on projects involving data science and analytics. If you are completely new to the subject, this course offers a non-technical grounding covering basic and some advanced principles and techniques that will certainly help anyone trying to get their head around the wider field of data science. If you want to truly understand data science then at some point you are going to come up against the field of statistics and probability, which can certainly be baffling for newcomers, particularly if your formal education days ended some time ago and what you did learn about the subject at school is a dim memory. This course explains how the statistical approach is used to make sense of the information that’s everywhere in the world around us. ———— You might also be interetsted in Bernard Marr’s latest book: Artificial Intelligence in Practice: How 50 Successful Companies Used Artificial Intelligence to Solve Problems And his video outlining the 7 biggest technology trends of 2020:

#Forbes

0 notes

yeyijia2019pp · 5 years ago

Text

POST1: Artificial Intelligence

A) AI in Education

Artificial intelligence is defined as the study of “intelligent agents”, by Russell (1995). With the development of science and technology, AI technology is increasingly used in daily life. I studied the use of AI technology in education.

Bernard Marr (2017) points out that AI has already been applied to education primarily in some tools that help develop skills and testing systems. As I know, some schools in China are using AI technology to monitor students' mental state during class:

The cameras in classroom use a computer vision algorithm to calculate the students’ performance in each class, based on facial expressions and movements. On the negative side, some people commented on social media that such behavior violated human rights. On the contrary, some people express their support for this kind of behavior. They believe that the use of AI to monitor students can help teachers get feedback on the classroom effect, so that teachers can modify the teaching methods and improve the teaching quality.

Images from: https://user.guancha.cn/main/content?id=172868?s=fwrphbios&from=singlemessage&isappinstalled=0

In addition to the use of AI face recognition in education. There are other ways AI can be used in education. Alyssa Johnson (2019) outlines five ways AI is changing the education industry: Simplifying Administrative Tasks, Smart Content, Personalized Learning, Global Learning and New Efficiencies.

AI is simplifying administrative tasks. For example, there are a lot of exams using computers to automatically mark scores. Such as China's high school college entrance examination and the International English Language Testing System.

AI technology can provide us with smart content. Such as adaptive learning system try to be more adaptable by building a model of each student's goals, preferences, and knowledge. Then using that model throughout the interaction with the student to suit the student's needs. (Peter Brusilovsky and Christoph Peylo, 2003).

AI technology also provides us with great convenience in improving efficiency. For example, in our university campus, there are many buildings and classrooms, and students are easy to get lost in the first class. We can use the ‘Herts mobile ’ application on campus to find the route accurately. Besides, this application can also choose the nearest route for you according to your location. This helps save time and increase efficiency.As a student of digital media design, I believe that studying the development of AI in education will help provide inspiration for my future creation. The existing network offer a variety of support to students and teachers who participate in intensive online education. In order to help understand such systems and ideas, Brusilovsky (1996) suggests focusing on adaptive and intelligent technologies. Thus, for example, we can design web interfaces or mobile apps for adaptive systems that improve teaching efficiency and strengthen the connection between teachers and students. In addition, we can add AR, 3D modeling, interaction design and other skills we learned in university.

Bibliography:

Barber, D. (2012). ‘Bayesian reasoning and machine learning’. Cambridge University Press.

El Janati, S., Maach, A., & El Ghanami, D. (2018). ‘SMART Education Framework for Adaptation Content Presentation’. Procedia Computer Science, 127, 436-443.

Holmes, W., Bialik, M., & Fadel, C. (2019). ‘Artificial intelligence in education: Promises and implications for teaching and learning’. Center for Curriculum Redesign.

Johnson, A. (2019). ‘5 Ways AI Is Changing the Education Industry’. ELearning Industry. [online] available at:https://elearningindustry.com/ai-is-changing-the-education-industry-5-ways

Marr, B. (2018) ‘How Is AI Used In Education -- Real World Examples of Today and a Peek into the Future’. Forbes. [online] available at:

https://www.forbes.com/sites/bernardmarr/2018/07/25/how-is-ai-used-in-education-real-world-examples-of-today-and-a-peek-into-the-future/#20de2e46586e

Neller, T. W. (2017). ‘AI education: Machine learning resources’. AI Matters, 3(2).

Peter, B. and Christoph, P. (2003). ‘Adaptive and Intelligent Web-based Educational Systems’. HAL CCSD.

Russell, S. J., & Norvig, P. (1995). ‘Artificial intelligence: a modern approach’. Englewood Cliffs, NJ: Prentice-Hall.

Siegersma, K. R., Leiner, T., Chew, D. P., Appelman, Y., Hofstra, L., & Verjans, J. W. (2019). ‘Artificial intelligence in cardiovascular imaging: state of the art and implications for the imaging cardiologist’. Netherlands Heart Journal, 1-11.

B) Group work

(groupmembers: Tamya Hussain,Xuefan Hao,Jingjun Ma,Camila Silva-Lonardelli)

In class, members of our group made up a story by using the key words journalist, pandemic, get rich, Dubai, AI smart cities and year 2049:

In 2049, we are journalists. However, we're journalists on the dark net. So just like the bad guys on the show, we want to get rich in Dubai by creating some events.

First, we spread around the city a powder that causes people to get allergic and turn their eyes red.

Then we let out the fake news that the city was facing a new pandemic, and only one company (our fake medical company) had produced the cure. This condition is very similar to ordinary pinkeye, so we require all people to go through an AI gate before buying an antidote to distinguish people with powder allergies from those with real pinkeye.

Eventually we not only became rich by selling antidotes, but became local heroes by curing everyone.

0 notes

besteccbest · 6 years ago

Text

iOS 14: Apple Changing How It Develops Its Software After Buggy iOS 13 Debut

Apple is overhauling how it tests software after a swarm of bugs marred the latest iPhone and iPad operating systems, according to people familiar with the shift. Software chief Craig Federighi and lieutenants including Stacey Lysik announced the changes at a recent internal "kickoff" meeting with the company's software developers. The new approach calls for Apple's development teams to ensure that test versions, known as "daily builds," of future software updates disable unfinished or buggy features by default. Testers will then have the option to selectively enable those features, via a new internal process and settings menu dubbed Flags, allowing them to isolate the impact of each individual addition on the system.

When the company's iOS 13 was released alongside the iPhone 11 in September, iPhone owners and app developers were confronted with a litany of software glitches. Apps crashed or launched slowly. Cellular signal was inconsistent. There were user interface errors in apps like Messages, system-wide search issues and problems loading emails. Some new features, such as sharing file folders over iCloud and streaming music to multiple sets of AirPods, were either delayed or are still missing. This amounted to one of the most troubled and unpolished operating system updates in Apple's history.

"iOS 13 continues to destroy my morale," Marco Arment, a well known developer, wrote on Twitter. "Same," replied Jason Marr, co-creator of grocery list app AnyList. "Apple's really shown a lack of respect for both its developers and its customers with iOS 13."

The issues show how complex iPhones have become and how easily users can be disappointed by a company known for the smooth integration of hardware and software. Annual software updates timed for release with the latest iPhones are a critical way for Apple to add new capabilities and keep users from defecting to archrival Android. Refreshed operating systems also give developers more tools for app creation, catalyzing more revenue for Apple from its App Store.

Apple spokeswoman Trudy Muller declined to comment.

The new development process will help early internal iOS versions to be more usable, or "livable," in Apple parlance. Prior to iOS 14's development, some teams would add features every day that weren't fully tested, while other teams would contribute changes weekly. "Daily builds were like a recipe with lots of cooks adding ingredients," a person with knowledge of the process said.

Test software got so crammed with changes at different stages of development that the devices often became difficult to use. Because of this, some "testers would go days without a livable build, so they wouldn't really have a handle on what's working and not working," the person said. This defeated the main goal of the testing process as Apple engineers struggled to check how the operating system was reacting to many of the new features, leading to some of iOS 13's problems.

Apple measures and ranks the quality of its software using a scale of 1 to 100 that's based on what's known internally as a "white glove" test. Buggy releases might get a score in the low 60s whereas more stable software would be above 80. iOS 13 scored lower on that scale than the more polished iOS 12 that preceded it. Apple teams also assign green, yellow and red color codes to features to indicate their quality during development. A priority scale of 0 through 5, with 0 being a critical issue and 5 being minor, is used to determine the gravity of individual bugs.

The new strategy is already being applied to the development of iOS 14, codenamed "Azul" internally, ahead of its debut next year. Apple has also considered delaying some iOS 14 features until 2021 - in an update called "Azul +1" internally that will likely become known as iOS 15 externally - to give the company more time to focus on performance. Still, iOS 14 is expected to rival iOS 13 in the breadth of its new capabilities, the people familiar with Apple's plans said.

The testing shift will apply to all of Apple's operating systems, including iPadOS, watchOS, macOS and tvOS. The latest Mac computer operating system, macOS Catalina, has also manifested bugs such as incompatibility with many apps and missing messages in Mail. Some HomePod speakers, which run an iOS-based operating system, stopped working after a recent iOS 13 update, leading Apple to temporarily pull the upgrade. The latest Apple Watch and Apple TV updates, on the other hand, have gone more smoothly.

Apple executives hope that the overhauled testing approach will improve the quality of the company's software over the long term. But this isn't the first time that Apple engineers have heard this from management.

Last year, Apple delayed several iOS 12 features - including redesigns for CarPlay and the iPad home screen - specifically so it could focus on reliability and performance. At an all-hands meeting in January 2018, Federighi said the company had prioritized new features too much and should return to giving consumers the quality and stability that they wanted first.

Apple then established so-called Tiger Teams to address performance issues in specific parts of iOS. The company reassigned engineers from across the software division to focus on tasks such as speeding up app launch times, improving network connectivity and boosting battery life. When iOS 12 came out in the fall of 2018, it was a stable release that required just two updates in the first two months.

That success didn't carry over to this year. The initial version of iOS 13 was so buggy that Apple has had to rush out several patches. In the first two months of iOS 13, there have been eight updates, the most since 2012 when Federighi took over Apple's iOS software engineering group. The company is currently testing another new version, iOS 13.3, and there's already a follow-up in the works for the spring.

About a month before Apple's 2019 Worldwide Developers Conference in June, the company's software engineers started to realize that iOS 13, then known internally as "Yukon," wasn't performing as well as previous versions. Some people who worked on the project said development was a "mess."

By August, realizing that the initial iOS 13.0 set to ship with new iPhones a few weeks later wouldn't hit quality standards, Apple engineers decided to mostly abandon that work and focus on improving iOS 13.1, the first update. Apple privately considered iOS 13.1 the "actual public release" with a quality level matching iOS 12. The company expected only die-hard Apple fans to load iOS 13.0 onto their phones.

The timing of the iOS 13.1 update was moved up by a week to Sept. 24, compressing the time that iOS 13.0 was Apple's flagship OS release. New iPhones are so tightly integrated with Apple software that it would have been technically impossible to launch the iPhone 11 with iOS 12, and since 13.1 wasn't ready in time, Apple's only choice was to ship with 13.0 and update everyone to 13.1 as quickly as it could.

While the iOS 13 issues did upset iPhone owners, they still updated fairly quickly. As of mid-October, half of all Apple device users were running a version of iOS 13, according to Apple. That upgrade pace is still far ahead of Google's Android.

Once iOS 13.1 was released, Apple's software engineering division pivoted to iOS 13.2 with a quality goal of being better than iOS 12. This update has had fewer complaints than its predecessors in the iOS 13 family but did introduce a short-lived bug around apps closing in the background when they shouldn't.

"iOS 13 has felt like a super-messy release, something we haven't seen this bad since iOS 8 or so," Steve Troughton-Smith, a veteran developer of Apple apps, wrote on Twitter.

Welcom to batteriestore.co.uk!We are a manufacturer representative and wholesaler that specializes in Laptop Batteries and Laptop AC Adapters . We have tested Batteries from virtually every manufacturer on the market. Our commitment is to provide our customers with the price/performance available on the market.

Please read the product details carefully to ensure that the brand, model and compatibility of the laptop battery you purchased meets your needs. If you need help, please contact us. Email: [email protected].

0 notes