Search Results for author:
Found 4 papers, 3 papers with code
TemporalMaxer: Maximize Temporal Context with only Max Pooling for Temporal Action Localization
To this end, we introduce TemporalMaxer, which minimizes long-term temporal context modeling while maximizing information from the extracted video clip features with a basic, parameter-free, and local region operating max-pooling block.
Ranked #1 on
Temporal Action Localization
on MUSES
SimOn: A Simple Framework for Online Temporal Action Localization
In addition, the evaluation for Online Detection of Action Start (ODAS) demonstrates the effectiveness and robustness of our method in the online setting.
Improving Object Detection by Label Assignment Distillation
Using PAA-ResNet50 as a teacher, our LAD techniques can improve detectors PAA-ResNet101 and PAA-ResNeXt101 to $46 \rm AP$ and $47. 5\rm AP$ on the COCO test-dev set.
1st Place Solution for YouTubeVOS Challenge 2021:Video Instance Segmentation
Video Instance Segmentation (VIS) is a multi-task problem performing detection, segmentation, and tracking simultaneously.
Ranked #20 on
Video Instance Segmentation
on YouTube-VIS validation
