Spatial Temporal Transformers
The discussion dives into the advantages of using spatial temporal transformers over traditional vision transformers, highlighting their ability to capture temporal dynamics while efficiently processing spatial information. Ashley emphasizes the interactive nature of their model, Genie, which allows users to actively engage with the AI, creating a more dynamic feedback loop that enhances the overall experience. This approach may signal a shift towards more engaging forms of AI interaction in the future.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Ashley Edwards - Genie Paper (DeepMind/Runway)
Related Questions