ARUDet: active retrieval and uncertainty-aware detection for sports video object detection
Lijing Yu, Jin Gao, Xinqi Ji, Bo Liu · 2026
Read ARUDet: active retrieval and uncertainty-aware detection for sports video object detection by Lijing Yu, Jin Gao, Xinqi Ji, Bo Liu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13469373.txt.
Description
Lijing Yu built a sports video detector in 2026 that actively repairs damaged regions and models uncertain edges. Fast motion, heavy occlusion, and warping defeat passive feature fusion, which treats all regions alike, while fixed box fitting ignores edge ambiguity. Their ARUDet pairs an active temporal recall module with a doubt-aware fitting head.
The recall module names each region's defect type, queries it, and pulls matching past features from memory to restore only low-quality areas. The fitting head models edge distributions and tightens worst-case error, making aim more sure. Tests across sports datasets showed clear gains, supporting use in TV, tactics, and athlete training.
Catalog record
| CREATOR | Lijing Yu, Jin Gao, Xinqi Ji, Bo Liu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 77 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |