References & Further Reading

References

1. K. He, X. Zhang, S. Ren, and J. Sun, Deep Residual Learning for Image Recognition, CVPR, 2016

   The ResNet paper. Introduced skip connections enabling training of 152-layer networks and winning ImageNet 2015.

2. O. Ronneberger, P. Fischer, and T. Brox, U-Net: Convolutional Networks for Biomedical Image Segmentation, MICCAI, 2015

   The U-Net paper. Encoder-decoder with skip connections for pixel-level prediction tasks.

3. K. Zhang, W. Zuo, Y. Chen, D. Meng, and L. Zhang, Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising, IEEE TIP, 2017

   DnCNN: demonstrated that a plain CNN with residual learning and BatchNorm could match BM3D for image denoising.

4. K. Zhang, Y. Li, W. Zuo, L. Zhang, L. Van Gool, and R. Timofte, Plug-and-Play Image Restoration with Deep Denoiser Prior, IEEE TPAMI, 2021

   DRUNet: U-Net denoiser with noise level map input, used as a plug-and-play prior for inverse problems.

5. S. Ioffe and C. Szegedy, Batch Normalization: Accelerating Deep Network Training, ICML, 2015

   Introduced BatchNorm, enabling higher learning rates and faster training of deep networks.

Further Reading

• ConvNeXt: modernising CNNs with transformer ideas

  Liu et al., A ConvNet for the 2020s, CVPR 2022

  Shows that pure CNNs can match Vision Transformers with modern design choices.

• Efficient architectures: MobileNet, EfficientNet

  Tan & Le, EfficientNet, ICML 2019

  Compound scaling and efficient blocks for mobile deployment.

• CNN-based channel estimation

  Soltani et al., IEEE Comm. Letters, 2019

  Applies CNNs to OFDM channel estimation on the resource grid.