Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation

Ho Kei Cheng, Yu-Wing Tai, Chi-Keung Tang. Published in NeurIPS, 2021.

[Project Page] [Code & Data] [arXiv] [PDF]

Rethinks Space-Time Networks to build image correspondences and uses L2 similarity to encourage diversified voting -- simpler, better, and more efficient.


Modular Interactive Video Object Segmentation: Interaction-to-Mask, Propagation and Difference-Aware Fusion

Ho Kei Cheng, Yu-Wing Tai, Chi-Keung Tang. Published in CVPR, 2021.

[Project Page] [Code & Data] [arXiv] [PDF]

A more user-friendly and efficient paradigm of iVOS in which interactions and propagations are decoupled, with the user’s intention captured by a novel difference-aware fusion module.


CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement

Ho Kei Cheng*, Jihoon Chung*, Yu-Wing Tai, Chi-Keung Tang. Published in CVPR, 2020.

[Project Page] [Code & Data] [arXiv] [PDF]

Refines segmentations (up to 4K and beyond) in a class-agnostic manner without using any high-resolution training data through a set of carefully designed cascade operations.