mirrors/mmselfsup
1
0
Fork 0
mirror of https://github.com/open-mmlab/mmselfsup.git synced 2025-06-03 14:59:38 +08:00
Code Issues Packages Projects Releases Wiki Activity

update modelzoo

Browse Source
This commit is contained in:
xieenze 2020-08-12 16:06:20 +08:00
parent 04484d1b5d
commit ab93dbf33d
1 changed files with 11 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
docs/MODEL_ZOO.md
View File

@ -44,3 +44,14 @@ Coming soon.
* Please follow [here](GETTING_STARTED.md#voc0712--coco17-object-detection) to run the evaluation.
<table><thead><tr><th rowspan="2">Method</th><th rowspan="2">Config</th><th rowspan="2">Remarks</th><th colspan="3">VOC07+12</th></tr><tr><td>AP50</td><td>AP</td><td>AP75</td></tr></thead><tbody><tr><td><a href="https://github.com/pytorch/vision/blob/master/torchvision/models/resnet.py" target="_blank" rel="noopener noreferrer">ImageNet</a></td><td>-</td><td>torchvision</td><td>81.58</td><td>54.19</td><td>59.80</td></tr><tr><td>Random</td><td>-</td><td>kaiming</td><td>59.02</td><td>32.83</td><td>31.60</td></tr><tr><td><a href="https://www.cv-foundation.org/openaccess/content_iccv_2015/papers/Doersch_Unsupervised_Visual_Representation_ICCV_2015_paper.pdf" target="_blank" rel="noopener noreferrer">Relative-Loc</a></td><td>selfsup/relative_loc/r50.py</td><td>default</td><td></td><td></td><td></td></tr><tr><td><a href="https://arxiv.org/abs/1803.07728" target="_blank" rel="noopener noreferrer">Rotation-Pred</a></td><td>selfsup/rotation_pred/r50.py</td><td>default</td><td>80.91</td><td>55.52</td><td>61.39</td></tr><tr><td><a href="https://arxiv.org/abs/1805.01978" target="_blank" rel="noopener noreferrer">NPID</a></td><td>selfsup/npid/r50.py</td><td>default</td><td>80.03</td><td>54.11</td><td>59.50</td></tr><tr><td><a href="https://arxiv.org/abs/1911.05722" target="_blank" rel="noopener noreferrer">MoCo</a></td><td>selfsup/moco/r50_v1.py</td><td>default</td><td>81.38</td><td>55.95</td><td>62.23</td></tr><tr><td><a href="https://arxiv.org/abs/2003.04297" target="_blank" rel="noopener noreferrer">MoCo v2</a></td><td>selfsup/moco/r50_v2.py</td><td>default</td><td>81.96</td><td>56.63</td><td>62.90</td></tr><tr><td><a href="https://arxiv.org/abs/2002.05709" target="_blank" rel="noopener noreferrer">SimCLR</a></td><td>selfsup/simclr/r50_bs256_ep200.py</td><td>default</td><td>79.41</td><td>51.54</td><td>55.63</td></tr><tr><td><a href="https://arxiv.org/abs/2006.07733" target="_blank" rel="noopener noreferrer">BYOL</a></td><td>selfsup/byol/r50.py</td><td>default</td><td></td><td></td><td></td></tr></tbody></table>
### COCO2017 Object Detection
**Note**
* This benchmark follows the evluation protocols set up by MoCo.
* Config: `benchmarks/detection/configs/coco_R_50_C4_2x_moco.yaml`.
* Please follow [here](GETTING_STARTED.md#voc0712--coco17-object-detection) to run the evaluation.
<table><thead><tr><th rowspan="2">Method</th><th rowspan="2">Config</th><th rowspan="2">Remarks</th><th colspan="6">COCO2017</th></tr><tr><td>AP50(Box)</td><td>AP(Box)</td><td>AP75(Box)</td><td>AP50(Mask)</td><td>AP(Mask)</td><td>AP75(Mask)</td></tr></thead><tbody><tr><td><a href="https://github.com/pytorch/vision/blob/master/torchvision/models/resnet.py" target="_blank" rel="noopener noreferrer">ImageNet</a></td><td>-</td><td>torchvision</td><td>59.9</td><td>40.0</td><td>43.1</td><td>56.5</td><td>34.7</td><td>36.9</td></tr><tr><td>Random</td><td>-</td><td>kaiming</td><td>54.6</td><td>35.6</td><td>38.2</td><td>51.5</td><td>31.4</td><td>33.5</td></tr><tr><td><a href="https://www.cv-foundation.org/openaccess/content_iccv_2015/papers/Doersch_Unsupervised_Visual_Representation_ICCV_2015_paper.pdf" target="_blank" rel="noopener noreferrer">Relative-Loc</a></td><td>selfsup/relative_loc/r50.py</td><td>default</td><td>59.6</td><td>40.0</td><td>43.5</td><td>56.5</td><td>35.0</td><td>37.3</td></tr><tr><td><a href="https://arxiv.org/abs/1803.07728" target="_blank" rel="noopener noreferrer">Rotation-Pred</a></td><td>selfsup/rotation_pred/r50.py</td><td>default</td><td>59.3</td><td>40.0</td><td>43.6</td><td>56.0</td><td>34.9</td><td>37.4</td></tr><tr><td><a href="https://arxiv.org/abs/1805.01978" target="_blank" rel="noopener noreferrer">NPID</a></td><td>selfsup/npid/r50.py</td><td>default</td><td>59.0</td><td>39.4</td><td>42.8</td><td>55.9</td><td>34.5</td><td>36.6</td></tr><tr><td><a href="https://arxiv.org/abs/1911.05722" target="_blank" rel="noopener noreferrer">MoCo</a></td><td>selfsup/moco/r50_v1.py</td><td>default</td><td>60.5</td><td>40.9</td><td>44.2</td><td>57.1</td><td>35.5</td><td>37.7</td></tr><tr><td><a href="https://arxiv.org/abs/2003.04297">MoCo v2</a></td><td>selfsup/moco/r50_v2.py</td><td>default</td><td>60.7</td><td>40.9</td><td>44.2</td><td>57.2</td><td>35.5</td><td>37.9</td></tr><tr><td><a href="https://arxiv.org/abs/2002.05709">SimCLR</a></td><td>selfsup/simclr/r50_bs256_ep200.py</td><td>default</td><td>59.1</td><td>39.6</td><td>42.9</td><td>55.9</td><td>34.6</td><td>37.1</td></tr><tr><td><a href="https://arxiv.org/abs/2006.07733">BYOL</a></td><td>selfsup/byol/r50.py</td><td>default</td><td></td><td></td><td></td><td></td><td></td><td></td></tr></tbody></table>