Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Vision language models are blind

vlmsareblind.github.io

Post reply on HN
Print view View original

181–190 of 202 posts

Re: Vision language models are blind

#181

Post by aezart »

drodgers wrote: post #144
cs702 wrote: post #7

Wow, that is embarrassingly bad performance for current SOTA models (GPT-4o, Gemini-1.5 Pro, Sonnet-3, Sonnet-3.5), which are advertised and sold as being able to understand images, e.g., for guiding the blind or tutoring children in geometry! The tasks at which they fail are ridiculously simple for human beings, including, for example: * counting the number of times two lines intersect; * detecting whether two circl…

I can't help but read comments like this as: "My talking dog always makes mistakes on calculus problems: how embarrassingly bad!" Has the expectation treadmill really advanced so quickly that sub-human performance on any category of problems is now an embarrassment?

To me I guess it suggests that these models are not using the correct approach. We keep finding new types of tasks the models are bad at, then the next model fixes those issues because those specific tasks are added to the training set. But that approach never results in a generalized problem solving ability, just an ability to solve all the problems we've thought of so far.

Re: Vision language models are blind

#182

Post by foldr »

TeMPOraL wrote: post #165
cs702 wrote: post #13

Earlier quoted context omitted.

It is embarrassingly, shockingly bad, because these models are advertised and sold as being capable of understanding images. Evidently, all these models still fall short.

They're hardly being advertised or sold on that premise. They advertise and sell themselves, because people try them out and find out they work , and tell their friends and/or audiences. ChatGPT is probably the single biggest bona-fide organic marketing success story in recorded history.

This is fantastic news for software engineers. Turns out that all those execs who've decided to incorporate AI into their product strategy have already tried it out and ensured that it will actually work.

Re: Vision language models are blind

#183

Post by maxlamb »

sushid wrote: post #154
energy123 wrote: post #134

Earlier quoted context omitted.

GPT-4o is very good at some visual tasks like optical character recognition. So the selective blindness might just be what you say here -- all of its capacity is dedicated to minimizing loss on a few narrow tasks that had the most training data (like OCR). So it's not necessarily an inherent failure of the architecture to generalize, it could just be a capacity issue that will naturally be resolved with more scale.

Is that not just traditional OCR applied on top of LLM?

No it’s not, it’s a multimodal transformer model.

Re: Vision language models are blind

#184

Post by slashdave »

akavi wrote: post #45

Speaking as someone with only a tenuous grasp of how VLMs work, this naïvely feels like a place where the "embodiement" folks might have a point: Humans have the ability to "refine" their perception of an image iteratively, focusing in on areas of interest, while VLMs have to process the entire image at the same level of fidelity. I'm curious if there'd be a way to emulate this (have the visual tokens be low fidelity…

These models have learned to focus on specific portions of an image (after all, this is the stated purpose of a transformer).

Re: Vision language models are blind

#185

Post by slashdave »

poikroequ wrote: post #147

It's ironic, they fail these seemingly simple tests that are trivial even for a child to solve. Yet, I used Gemini to read a postcard containing handwritten Russian cursive text with lots of visual noise (postmarks and whatnot). It was able to read the text and translate it into English. I didn't even need to tell it the text is Russian. On the one hand, it's incredible what these LLMs are capable of. On the other ha…

Simple for a child, yes. Because we have evolved our vision to recognize patterns like this, because they are important for survival. Reading Russian is not.

From an algorithmic point of view, these vision tasks are actually quite difficult to explicitly program.

Re: Vision language models are blind

#186

Post by slashdave »

jordan_bonecut wrote: post #117
mglz wrote: post #28

I tought some Computational Geometry courses and efficiently computing the intersections of N line segments is not as straightforward as you might initially think. Since somewhere some computation must be done to recognize this and LLMs are not specifically trained for this task, it's not suprising they struggle. In general, basic geometry seems under-explored by learning.

Yes, but so is telling if a photo contains a dog or understanding sentiment in a paragraph of text. Complexity isn't quite the issue, I think it is that there is a distinction between the type of reasoning which these models have learnt and that which is necessary for concrete mathematical reasoning.

The models do not reason. They have learned associations, because these associations have appeared in their training sets.

Re: Vision language models are blind

#187

Post by gnutrino »

simonw wrote: post #105
gnutrino wrote: post #76

My guess is that the systems are running image recognition models, and maybe OCR on images, and then just piping that data as tokens into an LLM. So you are only ever going to get results as good as existing images models with the results filtered through an LLM. To me, this is only interesting if compared with results of image recognition models that can already answer these types of questions (if they even exist, I…

That's not how they work. The original GPT-4 paper has some detail: https://cdn.openai.com/papers/gpt-4.pdf Or read up on PaliGemma: https://github.com/google-research/big_vision/blob/main/big_...

Thanks, I'll read up on this.

Re: Vision language models are blind

#188

Post by gnutrino »

Eisenstein wrote: post #90
gnutrino wrote: post #76

My guess is that the systems are running image recognition models, and maybe OCR on images, and then just piping that data as tokens into an LLM. So you are only ever going to get results as good as existing images models with the results filtered through an LLM. To me, this is only interesting if compared with results of image recognition models that can already answer these types of questions (if they even exist, I…

> My guess is that the systems are running image recognition models Your guess is incorrect. Look up CLIP, BLIP, and SigLip for an idea of how they work.

Will do, thank you.

Re: Vision language models are blind

#189

Post by dartos »

brookst wrote: post #177
SomaticPirate wrote: post #18

Earlier quoted context omitted.

There are quite a few "ai apologists" in the comments but I think the title is fair when these models are marketed towards low vision people ("Be my eyes" https://www.youtube.com/watch?v=Zq710AKC1gg ) as the equivalent to human vision. These models are implied to be human level equivalents when they are not. This paper demonstrates that there are still some major gaps where simple problems confound the models in unex…

I don’t see Be My Eyes or other similar efforts as “implied” to be equivalent to humans at all. They’re just new tools which can be very useful for some people. “These new tools aren’t perfect” is the dog bites man story of technology. It’s certainly true, but it’s no different than GPS (“family drives car off cliff because GPS said to”).

Based take

Re: Vision language models are blind

#190

Post by ben_w »

foldr wrote: post #182
TeMPOraL wrote: post #165

Earlier quoted context omitted.

They're hardly being advertised or sold on that premise. They advertise and sell themselves, because people try them out and find out they work , and tell their friends and/or audiences. ChatGPT is probably the single biggest bona-fide organic marketing success story in recorded history.

This is fantastic news for software engineers. Turns out that all those execs who've decided to incorporate AI into their product strategy have already tried it out and ensured that it will actually work.

> Turns out that all those execs who've decided to incorporate AI into their product strategy have already tried it out and ensured that it will actually work.

The 2-4-6 game comes to mind. They may well have verified the AI will work, but it's hard to learn the skill of thinking about how to falsify a belief.

Post reply on HN