Video Instance Segmentation 2019: A Winning Approach for Combined Detection, Segmentation, Classification and Tracking.
By Jonathon Luiten, Philip H. S. Torr, Bastian Leibe. 2019
27 citations recorded by the source corpus.
Video Instance Segmentation (VIS) is the task of localizing all objects in a video, segmenting them, tracking them throughout the video and classifying them into a set of predefined classes. In this work, divide VIS into these four parts: detection, segmentation, tracking and classification. We then develop algorithms for performing each of these four sub tasks individually, and combine these into a complete solution for VIS. Our solution is an adaptation of UnOVOST, the current best performing algorithm for Unsupervised Video Object Segmentation, to this VIS task. We benchmark our algorithm on the 2019 YouTube-VIS Challenge, where we obtain first place with an mAP score of 46.7%.
Subfields: Computer Vision and Pattern Recognition
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Alan L. Yuille
- Philip H. S. Torr
- Ming-Hsuan Yang
- Xiaogang Wang
- Dacheng Tao
- Chunhua Shen
- Jitendra Malik
Citations in the corpus
- Occluded Video Instance Segmentation: A Benchmark (2021, background): “Since the release of the YouTube-VIS dataset, video instance segmentation has attracted great attention in the computer vision community, arising a series of algorithms recently [49, 4, 1, 2, 27, 30, 41, 11].”
- Classifying, Segmenting, and Tracking Object Instances in Video with Mask Propagation (2019, methodology): “Our method achieves the best reported accuracy on the YouTube-VIS dataset [42], outperforming the ICCV 2019 challenge winner [28] despite being much simpler and using significantly less labeled data (1000x fewer images and 10x fewer bounding boxes).”
- HOTA: A Higher Order Metric for Evaluating Multi-object Tracking (2020, background): “One of these, as discussed in [49], is that by producing many different predictions with low confidence scores, it greatly increases the chances of obtaining a correct trajectory match and thus improving the score.”