Skip-Convolutions for Efficient Video Processing with Amir Habibian - #496
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
00:00
Optimizing Video Processing with AI
This chapter explores advancements in artificial intelligence focused on efficient video processing and generative models for video compression. The speaker shares their journey in AI, highlighting innovations like skip convolutions and neural network processors that leverage data redundancy. The discussion contrasts traditional methods with modern approaches that capitalize on temporal correlations in video, aiming to enhance computational efficiency in tasks such as human pose estimation.
Transcript
Play full episode
