Revisiting Class Imbalance

Reported on 6 benchmarks across 2 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision3 results

Semi-Supervised Object DetectiononCOCO 100% labeled data
mAP· 2023-06-04
44
best: 55.2 (MixPL)
Revisiting Class Imbalance for End-to-end Semi-Supervised Object Detection arXiv:2306.02268
Semi-Supervised Object DetectiononCOCO 10% labeled data
mAP· 2023-06-04
37.4
best: 44.6 (MixPL)
Revisiting Class Imbalance for End-to-end Semi-Supervised Object Detection arXiv:2306.02268
Semi-Supervised Object DetectiononCOCO 5% labeled data
mAP· 2023-06-04
32.21
best: 40.1 (Semi-DETR)
Revisiting Class Imbalance for End-to-end Semi-Supervised Object Detection arXiv:2306.02268

Methodology3 results

2D Object DetectiononCOCO 100% labeled data
mAP· 2023-06-04
44
best: 55.2 (MixPL)
Revisiting Class Imbalance for End-to-end Semi-Supervised Object Detection arXiv:2306.02268
2D Object DetectiononCOCO 10% labeled data
mAP· 2023-06-04
37.4
best: 44.6 (MixPL)
Revisiting Class Imbalance for End-to-end Semi-Supervised Object Detection arXiv:2306.02268
2D Object DetectiononCOCO 5% labeled data
mAP· 2023-06-04
32.21
best: 40.1 (Semi-DETR)
Revisiting Class Imbalance for End-to-end Semi-Supervised Object Detection arXiv:2306.02268