Shortcuts

DenseCL

Abstract

To date, most existing self-supervised learning methods are designed and optimized for image classification. These pre-trained models can be sub-optimal for dense prediction tasks due to the discrepancy between image-level prediction and pixel-level prediction. To fill this gap, we aim to design an effective, dense self-supervised learning method that directly works at the level of pixels (or local features) by taking into account the correspondence between local features. We present dense contrastive learning (DenseCL), which implements self-supervised learning by optimizing a pairwise contrastive (dis)similarity loss at the pixel level between two views of input images.

Models and Benchmarks

In this page, we provide benchmarks as much as possible to evaluate our pre-trained models. If not mentioned, all models are pre-trained on ImageNet-1k dataset.

Classification

The classification benchmarks includes 4 downstream task datasets, VOC, ImageNet, iNaturalist2018 and Places205. If not specified, the results are Top-1 (%).

VOC SVM / Low-shot SVM

The Best Layer indicates that the best results are obtained from which layers feature map. For example, if the Best Layer is feature3, its best result is obtained from the second stage of ResNet (1 for stem layer, 2-5 for 4 stage layers).

Besides, k=1 to 96 indicates the hyper-parameter of Low-shot SVM.

Self-Supervised Config Best Layer SVM k=1 k=2 k=4 k=8 k=16 k=32 k=64 k=96
resnet50_8xb32-coslr-200e feature5 82.5 42.68 50.64 61.74 68.17 72.99 76.07 79.19 80.55

ImageNet Linear Evaluation

The Feature1 - Feature5 don’t have the GlobalAveragePooling, the feature map is pooled to the specific dimensions and then follows a Linear layer to do the classification. Please refer to resnet50_mhead_linear-8xb32-steplr-90e_in1k for details of config.

Self-Supervised Config Feature1 Feature2 Feature3 Feature4 Feature5
resnet50_8xb32-coslr-200e 15.86 35.47 49.46 64.06 62.95
Algorithm Backbone Epoch Batch Size Results (Top-1 %) Links
Linear Eval Fine-tuning Pretrain Linear Eval Fine-tuning
DenseCL ResNet50 200 256 63.5 / config | model | log config | model | log /

Places205 Linear Evaluation

The Feature1 - Feature5 don’t have the GlobalAveragePooling, the feature map is pooled to the specific dimensions and then follows a Linear layer to do the classification. Please refer to resnet50_mhead_8xb32-steplr-28e_places205.py for details of config.

Self-Supervised Config Feature1 Feature2 Feature3 Feature4 Feature5
resnet50_8xb32-coslr-200e 21.32 36.20 43.97 51.04 50.45

ImageNet Nearest-Neighbor Classification

The results are obtained from the features after GlobalAveragePooling. Here, k=10 to 200 indicates different number of nearest neighbors.

Self-Supervised Config k=10 k=20 k=100 k=200
resnet50_8xb32-coslr-200e 48.2 48.5 46.8 45.6

Detection

The detection benchmarks includes 2 downstream task datasets, Pascal VOC 2007 + 2012 and COCO2017. This benchmark follows the evluation protocols set up by MoCo.

Pascal VOC 2007 + 2012

Please refer to config for details.

Self-Supervised Config AP50
resnet50_8xb32-coslr-200e 82.14

COCO2017

Please refer to config for details.

Self-Supervised Config mAP(Box) AP50(Box) AP75(Box) mAP(Mask) AP50(Mask) AP75(Mask)
resnet50_8xb32-coslr-200e

Segmentation

The segmentation benchmarks includes 2 downstream task datasets, Cityscapes and Pascal VOC 2012 + Aug. It follows the evluation protocols set up by MMSegmentation.

Pascal VOC 2012 + Aug

Please refer to config for details.

Self-Supervised Config mIOU
resnet50_8xb32-coslr-200e 69.47

Citation

@inproceedings{wang2021dense,
  title={Dense contrastive learning for self-supervised visual pre-training},
  author={Wang, Xinlong and Zhang, Rufeng and Shen, Chunhua and Kong, Tao and Li, Lei},
  booktitle={CVPR},
  year={2021}
}
Read the Docs v: dev-1.x
Versions
latest
stable
1.x
dev-1.x
dev
Downloads
pdf
html
epub
On Read the Docs
Project Home
Builds

Free document hosting provided by Read the Docs.