NLP vs. Computer Vision
Ishan discusses the challenges of prediction in natural language processing compared to computer vision, highlighting how the structured nature of language allows for better contextual understanding. While both fields face significant hurdles in achieving human-level performance, he suggests that advancements in NLP may come more easily due to its finite set of words and contexts. The conversation delves into the complexities of translating visual information, emphasizing the inherent difficulties in capturing and processing images compared to the relatively straightforward encoding of language.In this clip
From this podcast
Lex Fridman Podcast
Ishan Misra: Self-Supervised Deep Learning in Computer Vision | Lex Fridman Podcast #206
Related Questions