Generating Masks from Boxes by Mining Spatio-Temporal Consistencies in Videos
This work addresses the data scarcity problem for video object segmentation and tracking by leveraging more readily available bounding box annotations, which is significant for researchers and practitioners in computer vision.
The authors developed a method to generate segmentation masks from bounding box annotations in videos by mining spatio-temporal consistencies. This approach enabled the creation of accurate masks for large-scale tracking datasets, leading to state-of-the-art results in video object segmentation and tracking.
Segmenting objects in videos is a fundamental computer vision task. The current deep learning based paradigm offers a powerful, but data-hungry solution. However, current datasets are limited by the cost and human effort of annotating object masks in videos. This effectively limits the performance and generalization capabilities of existing video segmentation methods. To address this issue, we explore weaker form of bounding box annotations. We introduce a method for generating segmentation masks from per-frame bounding box annotations in videos. To this end, we propose a spatio-temporal aggregation module that effectively mines consistencies in the object and background appearance across multiple frames. We use our resulting accurate masks for weakly supervised training of video object segmentation (VOS) networks. We generate segmentation masks for large scale tracking datasets, using only their bounding box annotations. The additional data provides substantially better generalization performance leading to state-of-the-art results in both the VOS and more challenging tracking domain.