MCAR (ResNet101, 448x448)

Reported on 3 benchmarks across 1 task · 1 paper · 1 SOTA

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

Methodology3 results

Multi-Label ClassificationonPASCAL VOC 2012
mAP· 2020-07-03
94.3
best: 96.2 (Q2L-TResL(448 resolution))
SOTA
Learning to Discover Multi-Class Attentional Regions for Multi-Label Image Recognition arXiv:2007.01755
Multi-Label ClassificationonMS-COCO
mAP· 2020-07-03
83.8
best: 93.54 (ADDS(ViT-L-336, resolution 1344))
Learning to Discover Multi-Class Attentional Regions for Multi-Label Image Recognition arXiv:2007.01755
Multi-Label ClassificationonPASCAL VOC 2007
mAP· 2020-07-03
94.8
best: 97.3 (Q2L-CvT(ImageNet-21K pretrained, resolution 384))
Learning to Discover Multi-Class Attentional Regions for Multi-Label Image Recognition arXiv:2007.01755