AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
3min chapter
Weakly Supervised Causal Representation Learning with Johann Brehmer - #605
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Identifying Causal Variables in Video Compression
The paper is basically mostly one theorem and the proof that accompanies it. But we also have a practical implementation. And essentially it consists of a variational auto encoder. So there's an encoder that takes as input pixels and it outputs some latent variables. These latent variables are the causal variables of the system. Then there's a decoder that never said back to the pixel space. We train this on these paired data sets of before and after the intervention images. And then we show in a series of experiments with some, we start with some very similar toy data sets as we work towards simple image data sets that this works in practice.
00:00
Transcript
Episode notes
Get the Snipd
podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any
moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share
& Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode