mmpretrain/configs/densecl

DenseCL

Dense contrastive learning for self-supervised visual pre-training

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.

How to use it?

Predict image

from mmpretrain import inference_model

predict = inference_model('resnet50_densecl-pre_8xb32-linear-steplr-100e_in1k', 'demo/bird.JPEG')
print(predict['pred_class'])
print(predict['pred_score'])

Use the model

import torch
from mmpretrain import get_model

model = get_model('densecl_resnet50_8xb32-coslr-200e_in1k', pretrained=True)
inputs = torch.rand(1, 3, 224, 224)
out = model(inputs)
print(type(out))
# To extract features.
feats = model.extract_feat(inputs)
print(type(feats))

Train/Test Command

Prepare your dataset according to the docs.

Train:

python tools/train.py configs/densecl/densecl_resnet50_8xb32-coslr-200e_in1k.py

Test:

python tools/test.py configs/densecl/benchmarks/resnet50_8xb32-linear-steplr-100e_in1k.py https://download.openmmlab.com/mmselfsup/1.x/densecl/densecl_resnet50_8xb32-coslr-200e_in1k/resnet50_linear-8xb32-steplr-100e_in1k/resnet50_linear-8xb32-steplr-100e_in1k_20220825-f0f0a579.pth

Models and results

Pretrained models

Model	Params (M)	Flops (G)	Config	Download
densecl_resnet50_8xb32-coslr-200e_in1k	N/A	N/A	config	model | log

Image Classification on ImageNet-1k

Model	Pretrain	Params (M)	Flops (G)	Top-1 (%)	Config	Download
resnet50_densecl-pre_8xb32-linear-steplr-100e_in1k	DENSECL	N/A	N/A	63.50	config	model | log

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}
}