Comments URL: https://news.ycombinator.com/item?id=47772216

Points: 1

# Comments: 0

I trained a, custom and small, neural network to detect when people look at your screen. Whenever more faces than your allowed threshold are looking at your screen an alert will show. The model being ran locally ensures all data stays on your device - no cloud processing.

I would love feedback on the application and any features you think I should implement!

Comments URL: https://news.ycombinator.com/item?id=47406434

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=46982448

Points: 2

# Comments: 0

Also, the ANE only allows some operators to be ran on it right? There's very little transparency/control on what can be offloaded to it and cannot which makes using it difficult.

The worst part of it is as you say we all accept it and no one talks about it.

Is there any recommended reading you'd suggest to look into this more and the impacts of it?

Also I did run a test of FP16 vs FP32 for a large matmul on the Apple GPU and the FP16 calculation was 1.28x faster so it makes sense that they'd go for FP16 as a default.

Also generally I think CoreML isn't the best. The best solution for ORT would probably be to introduce a pure MPS provider (https://github.com/microsoft/onnxruntime/issues/21271), but given they've already bought into CoreML the effort may not be worth the reward for the core team. Which fair enough as it's a pretty mammoth task

Comments URL: https://news.ycombinator.com/item?id=46350075

Points: 98

# Comments: 17

You can find more details at my site soon: https://ym2132.github.io

AI assisted programming isn't inherently bad, but it does allow contributions from people who may not understand what exactly they are contributing. PR Guard aims to stop this.

It works by:

- Passing the diff of a PR to an LLM - The LLM returns 3 questions which the author must answer - The LLM then reviews the answers and decides whether or not they show the author understands their code

The point is to relieve some pressure on reviewers AND to enable users of AI assisted programming to learn in a new and engaging way.

Comments URL: https://news.ycombinator.com/item?id=46028738

Points: 2

# Comments: 1

ha, thats probably why I noticed the EyesOff accuracy drops so much at longer ranges, I suppose two models would do better but atm battery drain is a big issue.

I'm not sure if it's important or not, but the app comes from my own problems working in public so I'm happy to continue working on it. I do want to train and deploy an optimised model, something much smaller.

Sounds great, once a POC get's built I'll let you know and can see about the clinical side.

Thanks for the tips! I'll be sure to post something and reach out if I get round to implementing such a model.

Do you remember the cost of Mech Turk? It was something I wanted to use for EyesOff but never could get around the cost aspect.

I need some time to process everything you said, but the EyesOff model has pretty low accuracy at the moment. I'm sure some of these tidbits of info could help to improve the model, although my data is pretty messy in comparison. I had thought of doing more gaze tracking work for my model, but at long ranges it just breaks down completely (in my experience, happy to stand corrected if you're worked on that too).

Regarding the baby screener, I see how this approach could be very useful. If I get the time, I'll look into it a bit more and see what I can come up with. I'll let you know once I get round to it.

Is there any research/papers on this type of autism diagnosis tools for babies?

To your last point, yes I agree. Even the task I setup the model for is relatively easy compared to proper gaze tracking, I just rely on large datasets.

I suppose you could do it in the way you say and then from that gather data to eventually build out another model.

I'll for sure look into this, appreciate the idea sharing!

They create this CNN for exactly this task, autism diagnosis in children. I suppose this model would work for babies too.

Edit: ah I see your point, in the paper they diagnose autism with eye contact, but your point is a task closer to what my model does. It could definelty be augmented for such a task, we’d just need to improve the accuracy. The only issue I see is sourcing training data might be tricky, unless I partner with some institution researching this. If you know of anyone in this field I’d be happy to speak with them.