Weakly Supervised Semantic Segmentation using Web-Crawled Videos
This work addresses the challenge of improving semantic segmentation accuracy for computer vision applications with minimal supervision, though it is incremental as it builds on existing weakly supervised techniques.
The paper tackles the problem of weakly supervised semantic segmentation, where models often focus only on discriminative parts, by using web-crawled videos to generate strong segmentation labels without extra human supervision, resulting in performance that substantially outperforms existing methods and is competitive with approaches using additional annotations.
We propose a novel algorithm for weakly supervised semantic segmentation based on image-level class labels only. In weakly supervised setting, it is commonly observed that trained model overly focuses on discriminative parts rather than the entire object area. Our goal is to overcome this limitation with no additional human intervention by retrieving videos relevant to target class labels from web repository, and generating segmentation labels from the retrieved videos to simulate strong supervision for semantic segmentation. During this process, we take advantage of image classification with discriminative localization technique to reject false alarms in retrieved videos and identify relevant spatio-temporal volumes within retrieved videos. Although the entire procedure does not require any additional supervision, the segmentation annotations obtained from videos are sufficiently strong to learn a model for semantic segmentation. The proposed algorithm substantially outperforms existing methods based on the same level of supervision and is even as competitive as the approaches relying on extra annotations.