UVO Challenge on Video-based Open-World Segmentation 2021: 1st Place Solution

2021/10/22 by Yuming Du, W. Guo, Du, Yuming +5
Computer Science · #Advanced Image and Video Retrieval Techniques #Advanced Neural Network Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Image and Object Detection Techniques

paper · pdf · doi:10.48550/arxiv.2110.11661

openalex publication_date 2021/10/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this report, we introduce our (pretty straightforard) two-step "detect-then-match" video instance segmentation method. The first step performs instance segmentation for each frame to get a large number of instance mask proposals. The second step is to do inter-frame instance mask matching with the help of optical flow. We demonstrate that with high quality mask proposals, a simple matching mechanism is good enough for tracking. Our approach achieves the first place in the UVO 2021 Video-based Open-World Segmentation Challenge.

Citations

• The Pascal Visual Object Classes (VOC) Challenge
• ShapeNet: An Information-Rich 3D Model Repository
• MMDetection: Open MMLab Detection Toolbox and Benchmark
• Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution
• RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

Related

• 1st Place Solution for the UVO Challenge on Image-based Open-World Segmentation 2021
• Mask-Free Video Instance Segmentation
• SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation
• FlowCut: Unsupervised Video Instance Segmentation via Temporal Mask Matching
• Key Instance Selection for Unsupervised Video Object Segmentation