Mask2Former for Video Instance Segmentation

We find Mask2Former also achieves state-of-the-art performance on video instance segmentation without modifying the architecture, the loss or even the training pipeline. In this report, we show universal image segmentation architectures trivially generalize to video segmentation by directly predicting 3D segmentation volumes. Specifically, Mask2Former sets a new state-of-the-art of 60.4 AP on YouTubeVIS-2019 and 52.6 AP on YouTubeVIS-2021. We believe Mask2Former is also capable of handling video semantic and panoptic segmentation, given its versatility in image segmentation. We hope this will make state-of-the-art video segmentation research more accessible and bring more attention to designing universal image and video segmentation architectures.

PDF Abstract

Datasets


Task Dataset Model Metric Name Metric Value Global Rank Benchmark
Video Instance Segmentation YouTube-VIS validation Mask2Former (ResNet-101) mask AP 49.2 # 4
AP50 72.8 # 4
AP75 54.2 # 6
Video Instance Segmentation YouTube-VIS validation Mask2Former (ResNet-50) mask AP 46.4 # 8
AP50 68.0 # 8
AP75 50.0 # 9
Video Instance Segmentation YouTube-VIS validation Mask2Former (Swin-L) mask AP 60.4 # 1
AP50 84.4 # 1
AP75 67.0 # 1

Methods


No methods listed for this paper. Add relevant methods here