Spatio Temporal Activity Recognition
- 0 Collaborators
Perceiving dynamic actions could be a huge advance in how software makes sense of the world. For a computer, recognizing a cat or a duck in a still image is pretty clever. But a stiffer test for artificial intelligence will be understanding when the cat is riding a Roomba and chasing the duck around a kitchen. The aim of this project is to explore and leverage the temporal information to identify, localize and track objects over time. ...learn more
Project status: Under Development
Overview / Usage
Video understanding has an unfathomable potential if energized by AI. With the boom in deep learning, video understanding has delivered much better results than previous methods. Its use cases range from optimizing large-scale video annotation or automated thumbnails. It has widespread applications in robotics and is a field of active research (Google's time contrastive networks are really interesting).
Methodology / Approach
The aim of this project is to establish various benchmarks on large-scale datasets corresponding to video understanding research.