mmpretrain/configs/replknet

RepLKNet

Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs

Abstract

We revisit large kernel design in modern convolutional neural networks (CNNs). Inspired by recent advances in vision transformers (ViTs), in this paper, we demonstrate that using a few large convolutional kernels instead of a stack of small kernels could be a more powerful paradigm. We suggested five guidelines, e.g., applying re-parameterized large depth-wise convolutions, to design efficient highperformance large-kernel CNNs. Following the guidelines, we propose RepLKNet, a pure CNN architecture whose kernel size is as large as 31×31, in contrast to commonly used 3×3. RepLKNet greatly closes the performance gap between CNNs and ViTs, e.g., achieving comparable or superior results than Swin Transformer on ImageNet and a few typical downstream tasks, with lower latency. RepLKNet also shows nice scalability to big data and large models, obtaining 87.8% top-1 accuracy on ImageNet and 56.0% mIoU on ADE20K, which is very competitive among the state-of-the-arts with similar model sizes. Our study further reveals that, in contrast to small-kernel CNNs, large kernel CNNs have much larger effective receptive fields and higher shape bias rather than texture bias.

Results and models

ImageNet-1k

Model	Resolution	Pretrained Dataset	Params(M)	Flops(G)	Top-1 (%)	Top-5 (%)	Config	Download
RepLKNet-31B*	224x224	From Scratch	79.9（train) | 79.5 (deploy)	15.6 (train) | 15.4 (deploy)	83.48	96.57	config (train) | config (deploy)	model
RepLKNet-31B*	384x384	From Scratch	79.9（train) | 79.5 (deploy)	46.0 (train) | 45.3 (deploy)	84.84	97.34	config (train) | config (deploy)	model
RepLKNet-31B*	224x224	ImageNet-21K	79.9（train) | 79.5 (deploy)	15.6 (train) | 15.4 (deploy)	85.20	97.56	config (train) | config (deploy)	model
RepLKNet-31B*	384x384	ImageNet-21K	79.9（train) | 79.5 (deploy)	46.0 (train) | 45.3 (deploy)	85.99	97.75	config (train) | config (deploy)	model
RepLKNet-31L*	384x384	ImageNet-21K	172.7（train) | 172.0 (deploy)	97.2 (train) | 97.0 (deploy)	86.63	98.00	config (train) | config (deploy)	model
RepLKNet-XL*	320x320	MegData-73M	335.4（train) | 335.0 (deploy)	129.6 (train) | 129.0 (deploy)	87.57	98.39	config (train) | config (deploy)	model

Models with * are converted from the official repo. The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.

How to use

The checkpoints provided are all training-time models. Use the reparameterize tool to switch them to more efficient inference-time architecture, which not only has fewer parameters but also less calculations.

Use tool

Use provided tool to reparameterize the given model and save the checkpoint:

python tools/convert_models/reparameterize_model.py ${CFG_PATH} ${SRC_CKPT_PATH} ${TARGET_CKPT_PATH}

${CFG_PATH} is the config file, ${SRC_CKPT_PATH} is the source chenpoint file, ${TARGET_CKPT_PATH} is the target deploy weight file path.

To use reparameterized weights, the config file must switch to the deploy config files.

python tools/test.py ${Deploy_CFG} ${Deploy_Checkpoint} --metrics accuracy

In the code

Use backbone.switch_to_deploy() or classificer.backbone.switch_to_deploy() to switch to the deploy mode. For example:

from mmcls.models import build_backbone

backbone_cfg=dict(type='RepLKNet',arch='31B'),
backbone = build_backbone(backbone_cfg)
backbone.switch_to_deploy()

or

from mmcls.models import build_classifier

cfg = dict(
    type='ImageClassifier',
    backbone=dict(
        type='RepLKNet',
        arch='31B'),
    neck=dict(type='GlobalAveragePooling'),
    head=dict(
        type='LinearClsHead',
        num_classes=1000,
        in_channels=1024,
        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
        topk=(1, 5),
    ))

classifier = build_classifier(cfg)
classifier.backbone.switch_to_deploy()

Citation

@inproceedings{ding2022scaling,
  title={Scaling up your kernels to 31x31: Revisiting large kernel design in cnns},
  author={Ding, Xiaohan and Zhang, Xiangyu and Han, Jungong and Ding, Guiguang},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={11963--11975},
  year={2022}
}