Shortcuts

HorNet

HorNet: Efficient High-Order Spatial Interactions with Recursive Gated Convolutions

Abstract

Recent progress in vision Transformers exhibits great success in various tasks driven by the new spatial modeling mechanism based on dot-product self-attention. In this paper, we show that the key ingredients behind the vision Transformers, namely input-adaptive, long-range and high-order spatial interactions, can also be efficiently implemented with a convolution-based framework. We present the Recursive Gated Convolution (g nConv) that performs high-order spatial interactions with gated convolutions and recursive designs. The new operation is highly flexible and customizable, which is compatible with various variants of convolution and extends the two-order interactions in self-attention to arbitrary orders without introducing significant extra computation. g nConv can serve as a plug-and-play module to improve various vision Transformers and convolution-based models. Based on the operation, we construct a new family of generic vision backbones named HorNet. Extensive experiments on ImageNet classification, COCO object detection and ADE20K semantic segmentation show HorNet outperform Swin Transformers and ConvNeXt by a significant margin with similar overall architecture and training configurations. HorNet also shows favorable scalability to more training data and a larger model size. Apart from the effectiveness in visual encoders, we also show g nConv can be applied to task-specific decoders and consistently improve dense prediction performance with less computation. Our results demonstrate that g nConv can be a new basic module for visual modeling that effectively combines the merits of both vision Transformers and CNNs. Code is available at https://github.com/raoyongming/HorNet.

Results and models

ImageNet-1k

Model

Pretrain

resolution

Params(M)

Flops(G)

Top-1 (%)

Top-5 (%)

Config

Download

HorNet-T*

From scratch

224x224

22.41

3.98

82.84

96.24

config

model

HorNet-T-GF*

From scratch

224x224

22.99

3.9

82.98

96.38

config

model

HorNet-S*

From scratch

224x224

49.53

8.83

83.79

96.75

config

model

HorNet-S-GF*

From scratch

224x224

50.4

8.71

83.98

96.77

config

model

HorNet-B*

From scratch

224x224

87.26

15.59

84.24

96.94

config

model

HorNet-B-GF*

From scratch

224x224

88.42

15.42

84.32

96.95

config

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.

Pre-trained Models

The pre-trained models on ImageNet-21k are used to fine-tune on the downstream tasks.

Model

Pretrain

resolution

Params(M)

Flops(G)

Download

HorNet-L*

ImageNet-21k

224x224

194.54

34.83

model

HorNet-L-GF*

ImageNet-21k

224x224

196.29

34.58

model

HorNet-L-GF384*

ImageNet-21k

384x384

201.23

101.63

model

*Models with * are converted from the official repo.

Citation

@article{rao2022hornet,
  title={HorNet: Efficient High-Order Spatial Interactions with Recursive Gated Convolutions},
  author={Rao, Yongming and Zhao, Wenliang and Tang, Yansong and Zhou, Jie and Lim, Ser-Lam and Lu, Jiwen},
  journal={arXiv preprint arXiv:2207.14284},
  year={2022}
}
Read the Docs v: master
Versions
master
latest
1.x
dev-1.x
Downloads
html
epub
On Read the Docs
Project Home
Builds

Free document hosting provided by Read the Docs.