A summary for state-of-the-art image-classifying methods on CIFAR-10, CIFAR-100 and ImageNet-1K
| Method | Depth | Params | CIFAR-10 | CIFAR-100 |
|---|---|---|---|---|
| Deeply Supervised Net [1] | - | - | 7.97% | 34.57% |
| Highway Network [2] | - | - | 7.72% | 32.39% |
| FractalNet [3] | 21 | 38.6M | 5.22% | 23.30% |
| FractalNet (Dropout) [3] | 21 | 38.6M | 4.60% | 23.73% |
| ResNet-110 [4] | 110 | 1.7M | 6.43% | - |
| ResNet-1202 [4] | 1202 | 19.4M | 7.93% | - |
| preact-ResNet-164 [5] | 164 | 1.7M | 5.46% | 24.33% |
| preact-ResNet-1001 [5] | 1001 | 10.7M | 4.92% | 22.71% |
| WRN-28-10 [6] | 28 | 36.5M | 4.00% | 19.25% |
| WRN-28-10 (dropout) [6] | 28 | 36.5M | 3.89% | 18.85% |
| AttentionNet-92 [7] | 92 | 1.9M | 4.99% | 21.71% |
| AttentionNet-236 [7] | 236 | 5.1M | 4.14% | 21.16% |
| AttentionNet-452 [7] | 452 | 8.6M | 3.90% | 20.45% |
| ResNeXt-29, 8×64d [8] | 29 | 34.4M | 3.65% | 17.77% |
| ResNeXt-29, 16×64d [8] | 29 | 68.1M | 3.58% | 17.31% |
| DenseNet-BC (k = 12) [9] | 100 | 0.8M | 4.51% | 22.27% |
| DenseNet-BC (k = 24) [9] | 250 | 15.3M | 3.62% | 17.60% |
| DenseNet-BC (k = 40) [9] | 190 | 25.6M | 3.46% | 17.18% |
| PyramidNet (alpha=200)[10] | 272 | 26.0M | 3.31% | 16.35% |
[1] Lee, Chen-Yu, Saining Xie, Patrick Gallagher, Zhengyou Zhang, and Zhuowen Tu. "Deeply-supervised nets." In Artificial Intelligence and Statistics, pp. 562-570. 2015.
[2] Srivastava, Rupesh Kumar, Klaus Greff, and Jürgen Schmidhuber. "Highway networks." arXiv preprint arXiv:1505.00387 (2015).
[3] Larsson, Gustav, Michael Maire, and Gregory Shakhnarovich. "Fractalnet: Ultra-deep neural networks without residuals." arXiv preprint arXiv:1605.07648 (2016).
[4] He, Kaiming, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. "Deep residual learning for image recognition." In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770-778. 2016.
[5] He, Kaiming, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. "Identity mappings in deep residual networks." In European Conference on Computer Vision, pp. 630-645. Springer, Cham, 2016.
[6] Zagoruyko, Sergey, and Nikos Komodakis. "Wide residual networks." arXiv preprint arXiv:1605.07146 (2016).
[7] Wang, Fei, Mengqing Jiang, Chen Qian, Shuo Yang, Cheng Li, Honggang Zhang, Xiaogang Wang, and Xiaoou Tang. "Residual attention network for image classification." arXiv preprint arXiv:1704.06904 (2017).
[8] Xie, Saining, Ross Girshick, Piotr Dollár, Zhuowen Tu, and Kaiming He. "Aggregated residual transformations for deep neural networks." In Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on, pp. 5987-5995. IEEE, 2017.
[9] Huang, Gao, Zhuang Liu, Kilian Q. Weinberger, and Laurens van der Maaten. "Densely connected convolutional networks." In Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on, 2017.
[10] Han, Dongyoon, Jiwhan Kim, and Junmo Kim. "Deep pyramidal residual networks." arXiv preprint arXiv:1610.02915 (2016).