Res2Net: A New Multi-Scale Backbone Architecture

@article{Gao2021Res2NetAN,
  title={Res2Net: A New Multi-Scale Backbone Architecture},
  author={Shanghua Gao and Ming-Ming Cheng and Kai Zhao and Xinyu Zhang and Ming-Hsuan Yang and Philip H. S. Torr},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year={2021},
  volume={43},
  pages={652-662}
}
Representing features at multiple scales is of great importance for numerous vision tasks. [...] Key Method The proposed Res2Net block can be plugged into the state-of-the-art backbone CNN models, e.g., ResNet, ResNeXt, and DLA. We evaluate the Res2Net block on all these models and demonstrate consistent performance gains over baseline models on widely-used datasets, e.g., CIFAR-100 and ImageNet. Further ablation studies and experimental results on representative computer vision tasks, i.e., object detection…Expand
Image super-resolution via enhanced multi-scale residual network
Lightweight Attended Multi-Scale Residual Network for Single Image Super-Resolution
HS-ResNet: Hierarchical-Split Block on Convolutional Neural Network
Lightweight Image Super-Resolution via Weighted Multi-Scale Residual Network
CSPNet: A New Backbone that can Enhance Learning Capability of CNN
Hierarchically Aggregated Residual Transformation for Single Image Super Resolution
  • Zejiang Hou, S. Kung
  • Computer Science
  • 2020 25th International Conference on Pattern Recognition (ICPR)
  • 2021
MADNet: A Fast and Lightweight Network for Single-Image Super Resolution
PSConv: Squeezing Feature Pyramid into One Compact Poly-Scale Convolutional Layer
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 78 REFERENCES
Dual Path Networks
Squeeze-and-Excitation Networks
Feature Pyramid Networks for Object Detection
Aggregated Residual Transformations for Deep Neural Networks
Richer Convolutional Features for Edge Detection
Very Deep Convolutional Networks for Large-Scale Image Recognition
Rethinking the Inception Architecture for Computer Vision
Network In Network
Wide Residual Networks
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
...
1
2
3
4
5
...