DETR-ResNet50 with iRPE-K (300 epochs)

Reported on 5 benchmarks across 5 tasks · 1 paper

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

Methodology4 results

3DonCOCO minival
box AP· 2021-07-29
42.3
best: 66 (PE_spatial (DETA))
Rethinking and Improving Relative Position Encoding for Vision Transformer arXiv:2107.14222
2D ClassificationonCOCO minival
box AP· 2021-07-29
42.3
best: 66 (PE_spatial (DETA))
Rethinking and Improving Relative Position Encoding for Vision Transformer arXiv:2107.14222
2D Object DetectiononCOCO minival
box AP· 2021-07-29
42.3
best: 66 (PE_spatial (DETA))
Rethinking and Improving Relative Position Encoding for Vision Transformer arXiv:2107.14222
16konCOCO minival
box AP· 2021-07-29
42.3
best: 66 (PE_spatial (DETA))
Rethinking and Improving Relative Position Encoding for Vision Transformer arXiv:2107.14222

Computer Vision1 result

Object DetectiononCOCO minival
box AP· 2021-07-29
42.3
best: 66 (PE_spatial (DETA))
Rethinking and Improving Relative Position Encoding for Vision Transformer arXiv:2107.14222