CDLNet: Robust and Interpretable Denoising Through Deep Convolutional Dictionary Learning
This work addresses the need for robust and interpretable denoising methods in computer vision, though it is incremental as it builds on existing unrolled optimization networks.
The authors tackled the problem of image denoising by proposing CDLNet, an unrolled convolutional dictionary learning network, which outperforms state-of-the-art models at similar parameter counts and achieves near-perfect generalization on unseen noise levels.
Deep learning based methods hold state-of-the-art results in image denoising, but remain difficult to interpret due to their construction from poorly understood building blocks such as batch-normalization, residual learning, and feature domain processing. Unrolled optimization networks propose an interpretable alternative to constructing deep neural networks by deriving their architecture from classical iterative optimization methods, without use of tricks from the standard deep learning tool-box. So far, such methods have demonstrated performance close to that of state-of-the-art models while using their interpretable construction to achieve a comparably low learned parameter count. In this work, we propose an unrolled convolutional dictionary learning network (CDLNet) and demonstrate its competitive denoising performance in both low and high parameter count regimes. Specifically, we show that the proposed model outperforms the state-of-the-art denoising models when scaled to similar parameter count. In addition, we leverage the model's interpretable construction to propose an augmentation of the network's thresholds that enables state-of-the-art blind denoising performance and near-perfect generalization on noise-levels unseen during training.