Video Inference Challenges
Reza discusses the complexities of video inference, highlighting that it is more computationally intensive than training due to the continuous nature of processing video streams. He explains the need for a robust caching infrastructure to manage millions of models effectively while utilizing Kubernetes for resource allocation. The orchestration of these components resembles a symphony, where each part plays a crucial role in the overall functionality.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Video Object Detection at Scale with Reza Zadeh - #34
Related Questions
Can you explain more about how AI models are trained?
What is the process of training a machine learning model?
What's an example of something that was hard to do before Kubernetes that is now easier to do with Kubernetes, specifically in relation to the episode Racing the Playhead: Real-time Model Inference in a Video Streaming Environment // Brannon Dorsey // Coffee Sessions #98 and the clip Kubernetes Complexity?