Original dev here. Unable to replicate this on my end, try refreshing?

Interesting. The issue occurs because I have WebGL disabled, causing the createShader function you're using to throw an error. You can reproduce this by going to chrome://settings, disabling "Use hardware acceleration when available," refreshing the page, and then triggering the same error.

Sorry for duplicate post. Also this feature is enabled by default, but causes issues with several sites.

Original dev here. I tend to agree for this particular demo video as I'm reading a book and I don't blink in the original.

The model tries to copy the blinks of the original video so it's possible that in other conditions, you'd notice less of this.

Fun to see this feedback though, definitely something worth improving :)

I likewise find the “corrections” uncanny. It’s not just the one with the book.

BTW your main site is throwing an error. Probably want to edit since your post is growing.

https://www.sievedata.com/

Application error: a client-side exception has occurred (see the browser console for more information).

Original dev here. Unable to replicate this on my end, try refreshing?

Interesting. The issue occurs because I have WebGL disabled, causing the createShader function you're using to throw an error. You can reproduce this by going to chrome://settings, disabling "Use hardware acceleration when available," refreshing the page, and then triggering the same error.

Original dev here. That's right, NVIDIA has a version available which we reference in our blog.

https://www.sievedata.com/blog/eye-contact-correction-gaze-c...

Newer models have come out that allow the same thing to be done and control even more than the eyes.

See here: https://github.com/KwaiVGI/LivePortrait/blob/main/assets/doc...

For web-conferencing, local use is great so NVIDIA's tools are what we recommend in that case.

I appreciated using the Nvidia Tools for remapping webcam eye-contact until I was reviewing a recording and noticed that it changed my eye color. But it’s been a bit. Perhaps an undocumented feature that newer versions/models fixed.

This is an interesting idea. We are a little farther off from being able to do this but agree it would look really cool.

Original dev here. Agree this video looks like uncanny valley but it's likely because the lighting of the original video is off + I baggy eyes (I was sleep deprived).

Would recommend trying it on other videos, it is surprisingly good. Although there definitely are areas to improve.

Congrats on the launch! I work at Sieve (https://www.sievedata.com/). We do a bunch of stuff with AI and video. Excited to check this out :)

Hi Mokshith, we talked about releasing a project using the Sieve API and Revideo before! This is definitely something we still want to do. I'll reach out!

Hey folks! We just built a cost-effective, lightweight way to generate audiovisual summaries for videos.

* Process videos up to 12x faster than realtime * Costs <$0.01 / min of video * Combines visual and audial components

The goal here is not to build a single E2E model but something that could actually be used in production while preserving relatively high quality.

You can try it out yourself here: https://www.sievedata.com/functions/sieve/describe How we built it: https://www.sievedata.com/blog/describe-video-summary-beta-l... The code: https://github.com/sieve-community/describe

What was the orchestration system used for? Was it in the case where there were many models that needed to be run one after another. I know it's a huge problem in video processing to be able to increase speed a ton. My company Sieve (see profile) is building infrastructure specifically for running ML models on video which is why I'm curious.

It was built for AI driven container orchestration, configuration synthesis from high level constraints.

Yes ML workloads is particularly complex, because they have both batch oriented data flows (training), and service oriented data flows (inference). There aren't many systems that can adequately express both.

Yeah sorry about that! We'll add that soon.

This looks really cool. Are there ways to orchestrate jobs? Like having one notebooks output trigger another based on some logic? I'm imagining running a bunch of different deep learning models on separate notebooks, or running the same model on different chunks of a piece of data in parallel.

Yes, you should follow best practices and isolate each job to the smallest task possible (and then reuse components). We have this functionality in 2 flavors, you can define hooks as part of your pipelines(https://docs.ploomber.io/en/latest/api/spec.html#id2), in addition you can define this dependency as part of your jobs DAG (https://docs.ploomber.io/en/latest/get-started/basic-concept...), e.g get the data, clean it, train the model and test it.