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

151–160 of 202 posts

Re: Vision language models are blind

#151

Post by ec109685 »

vessenes wrote: post #35

A few comments below talk about how tokenizing images using stuff like CLIP de-facto yields blurry image descriptions, and so these are ‘blind’ by some definitions. Another angle of blurring not much discussed is that the images are rescaled down; different resolutions for different models. I wouldn’t be surprised if Sonnet 3.5 had a higher res base image it feeds in to the model. Either way, I would guess that we’ll…

At least for gpt 4o, it can create a facsimile of images that it still can’t analyze properly, so I think it’s more than just its “eyes” that are broken.

It clearly wasn’t trained on this task and suffers accordingly.

However, with chatgpt, it will create python to do the analysis and has better results.

Re: Vision language models are blind

#152

Post by zwnow »

Melomololotolo wrote: post #93
sweezyjeezy wrote: post #23

Earlier quoted context omitted.

The paper I linked should hopefully mark me out as far from an AI apologist, it's actually really bad news for GenAI if correct. All I mean to say is the clickbait conclusion and the evidence do not match up.

We have started the ara of ai. It really doesn't matter how good current llms are. They have been good enough to start this ara. And no it's not and never has been just llms. Look what Nvidia is doing with ml. Whisper huge advantage, segment anything again huge. Alpha fold 2 again huge. All the robot announcements -> huge I doubt we will reach agi just through llms. We will reach agi through multi modal, mix of exper…

We won't reach agi in our lifetimes.

Re: Vision language models are blind

#153

Post by necovek »

vikramkr wrote: post #70
FrenchDevRemote wrote: post #47

Earlier quoted context omitted.

> their vision is, at best, like that of a person with myopia seeing fine details as blurry It's not that far from reality, most models sees images in very low resolution/limited colors, so not so far from this description

They didn't test that claim at all though. Vision isn't some sort of 1D sliding scale with every vision condition lying along one axis. First of all myopia isn't 'seeing fine details as blurry' - it's nearsightedness - and whatever else this post tested it definitely didn't test depth perception. And second - inability to see fine details is a distinct/different thing from not being able to count intersections and th…

It seems to me they've brought up myopia only to make it more approachable to people how blurry something is, implying they believe models work with a blurry image just like a nearsighted person sees blurry images at a distance.

While myopia is common, it's not the best choice of analogy and "blurry vision" is probably clear enough.

Still, I'd only see it as a bad choice of analogy — I can't imagine anyone mistaking optical focus problems for static image processing problems — so in the usual HN recommendation, I'd treat their example in the most favourable sense.

Re: Vision language models are blind

#154

Post by sushid »

energy123 wrote: post #134
joelburget wrote: post #128

Vision Transformers do a shocking amount of compression in the tokenizer. In the [Chameleon paper]( https://arxiv.org/pdf/2405.09818 ) they say the tokenizer "encodes a 512 × 512 image into 1024 discrete tokens from a codebook of size 8192". That's 256 pixels per token (512 * 512 / 1024). If we assume that a pixel is 24 bits (3x 8 bit channels), this implies that they've compressed 256 * 24 = 6144 bits into 13 = (log…

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?

Re: Vision language models are blind

#155

Post by kenjackson »

SomaticPirate wrote: post #18
sweezyjeezy wrote: post #8

Entertaining, but I think the conclusion is way off. > their vision is, at best, like that of a person with myopia seeing fine details as blurry is a crazy thing to write in an abstract. Did they try to probe that hypothesis at all? I could (well actually I can't) share some examples from my job of GPT-4v doing some pretty difficult fine-grained visual tasks that invalidate this. Personally, I rate this paper [1], wh…

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…

Simple is a relative statement. There are vision problems where monkeys are far better than humans. Some may look at human vision and memory and think that we lack basic skills.

With AINwe are creating intelligence but with different strengths and weaknesses. I think we will continue to be surprised at how well they work on some problems and how poor they do at some “simple” ones.

Re: Vision language models are blind

#156

Post by energy123 »

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?

It's possible they have a software layer that does that. But I was assuming they don't, because the open source multimodal models don't.

Re: Vision language models are blind

#158

Post by lynx23 »

SomaticPirate wrote: post #18
sweezyjeezy wrote: post #8

Entertaining, but I think the conclusion is way off. > their vision is, at best, like that of a person with myopia seeing fine details as blurry is a crazy thing to write in an abstract. Did they try to probe that hypothesis at all? I could (well actually I can't) share some examples from my job of GPT-4v doing some pretty difficult fine-grained visual tasks that invalidate this. Personally, I rate this paper [1], wh…

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…

Be My Eyes user here. I disagree with your uninformed opinion. Be My Eyes is more often than not more useful then a human. And I am reporting from personal experience. What experience do you have?

Re: Vision language models are blind

#159

Post by TeMPOraL »

pixl97 wrote: post #54
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…

Yea, really if you look at human learning/seeing/acting there is a feedback loop that LLM for example isn't able to complete and train on. You see an object. First you have to learn how to control all your body functions to move toward it and grasp it. This teaches you about the 3 dimensional world and things like gravity. You may not know the terms, but it is baked in your learning model. After you get an object you…

Embodied is useful, but I think not necessary even if you need learning in a 3D environment. Synthesized embodiment should be enough. While in some cases[0] it may have problems with fidelity, simulating embodied experience in silico scales much better, and more importantly, we have control over time flow. Humans always learn in real-time, while with simulated embodiment, we could cram years of subjective-time experiences into a model in seconds, and then for novel scenarios, spend an hour per each second of subjective time running a high-fidelity physics simulation[1].

--

[0] - Like if you plugged a 3D game engine into the training loop.

[1] - Results of which we could hopefully reuse in training later. And yes, a simulation could itself be a recording of carefully executed experiment in real world.

Re: Vision language models are blind

#160

Post by TeMPOraL »

sweezyjeezy wrote: post #8

Entertaining, but I think the conclusion is way off. > their vision is, at best, like that of a person with myopia seeing fine details as blurry is a crazy thing to write in an abstract. Did they try to probe that hypothesis at all? I could (well actually I can't) share some examples from my job of GPT-4v doing some pretty difficult fine-grained visual tasks that invalidate this. Personally, I rate this paper [1], wh…

Entertaining is indeed the right word. Nice job identifying corner cases of models' visual processing; curiously, they're not far conceptually from some optical illusions that reliably trip humans up. But to call the models "blind" or imply their low performance in general? That's trivially invalidated by just taking your phone out and feeding a photo to ChatGPT app.

Like, seriously. One poster below whines about "AI apologists" and BeMyEyes, but again, it's all trivially testable with your phone and $20/month subscription. It works spectacularly well on real world tasks. Not perfectly, sure, but good enough to be useful in practice and better than alternatives (which often don't exist).

Post reply on HN