Long-Term Image Boundary Prediction
This addresses the challenge of long-term visual prediction for applications in robotics or video analysis, but it is incremental as it extends boundary estimation from observed to future frames.
The paper tackles the problem of predicting image boundaries for future unobserved frames in videos, requiring learning about boundary dynamics and motion patterns, and demonstrates results on natural and synthetic sequences, with fusion of RGB and boundary prediction improving RGB predictions.
Boundary estimation in images and videos has been a very active topic of research, and organizing visual information into boundaries and segments is believed to be a corner stone of visual perception. While prior work has focused on estimating boundaries for observed frames, our work aims at predicting boundaries of future unobserved frames. This requires our model to learn about the fate of boundaries and corresponding motion patterns -- including a notion of "intuitive physics". We experiment on natural video sequences along with synthetic sequences with deterministic physics-based and agent-based motions. While not being our primary goal, we also show that fusion of RGB and boundary prediction leads to improved RGB predictions.