Object Detection on MS COCO (train val)
32.87APOurs
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| OursImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 32.87 | 55.15 | 34.58 | 18.99 | 35.57 | 41.44 | |
| ENMSImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 32.56 | 54.84 | 34.25 | 18.53 | 35.42 | 40.96 | |
| DivProtoImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 32.53 | 54.77 | 34.24 | 18.57 | 35.34 | 41.44 | |
| EntropyImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 32.36 | 54.69 | 33.97 | 18.67 | 35.25 | 40.63 | |
| Learn LossImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 32.33 | 54.61 | 33.92 | 18.72 | 35.37 | 40.76 | |
| MIALImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 32.27 | 54.69 | 34.04 | 17.72 | 35.27 | 41.86 | |
| OursImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.99 | 54.18 | 33.51 | 18.09 | 34.39 | 40.54 | |
| DivProtoImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.86 | 54.08 | 33.41 | 18.24 | 34.58 | 40.56 | |
| ENMSImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.79 | 54.05 | 33.29 | 18.11 | 34.5 | 40.14 | |
| MIALImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.75 | 53.89 | 33.42 | 17.51 | 34.6 | 41.09 | |
| CDALImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.75 | 53.57 | 33.38 | 17.75 | 34.53 | 41.23 | |
| RandomImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.62 | 53.29 | 33.18 | 17.14 | 34.19 | 41.26 | |
| EntropyImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.47 | 53.59 | 32.95 | 18.01 | 34.16 | 39.76 | |
| Core-setImages budget=40%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.31 | 52.89 | 32.8 | 16.19 | 33.81 | 42.85 | |
| Learn LossImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.19 | 53.19 | 32.67 | 17.48 | 34.04 | 39.23 | |
| CDALImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 31.17 | 53.07 | 32.55 | 17.22 | 33.84 | 40.56 | |
| RandomImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.99 | 52.7 | 32.47 | 16.88 | 33.52 | 40.35 | |
| OursImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.9 | 53.08 | 32.01 | 17.56 | 33.44 | 39.1 | |
| ENMSImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.82 | 52.89 | 32.07 | 17.4 | 33.48 | 38.88 | |
| Core-setImages budget=35%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.69 | 52.25 | 31.97 | 15.96 | 33.17 | 41.76 | |
| DivProtoImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.64 | 52.64 | 31.77 | 17.21 | 33.37 | 38.92 | |
| EntropyImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.61 | 52.66 | 31.96 | 17.49 | 33.13 | 38.97 | |
| Learn LossImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.55 | 52.53 | 31.9 | 17.68 | 33.1 | 38.52 | |
| MIALImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.47 | 52.49 | 31.45 | 16.85 | 33.44 | 39.31 | |
| RandomImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.07 | 51.61 | 31.13 | 16.05 | 32.42 | 39.42 | |
| CDALImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 30.01 | 51.57 | 31.27 | 16.28 | 32.72 | 39.21 | |
| Core-setImages budget=30%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.9 | 51.37 | 31.12 | 15.6 | 32.19 | 40.65 | |
| OursImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.78 | 51.7 | 30.81 | 16.52 | 32.32 | 38 | |
| ENMSImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.76 | 51.64 | 30.78 | 16.72 | 32.12 | 38.03 | |
| DivProtoImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.73 | 51.45 | 30.84 | 16.62 | 32.32 | 38.08 | |
| Learn LossImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.42 | 51.08 | 30.38 | 16.35 | 31.96 | 37.81 | |
| MIALImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.39 | 51.21 | 30.36 | 15.96 | 31.87 | 38.68 | |
| EntropyImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.38 | 51.06 | 30.31 | 16.18 | 31.63 | 37.89 | |
| CDALImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 29.11 | 48.6 | 27.86 | 14.29 | 29.69 | 35.82 | |
| RandomImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 28.97 | 50.43 | 29.84 | 15.39 | 31.25 | 37.84 | |
| Core-setImages budget=25%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 28.75 | 50 | 29.56 | 14.59 | 30.98 | 38.79 | |
| RandomImages budget=20%, Backbone=ResNet-50, Detector=Faster R-CNN2022.04 | 27.57 | 48.81 | 28.09 | 14.53 | 29.87 | 36.12 |