H-OWAN: Multi-distorted Image Restoration with Tensor 1x1 Convolution
arXiv:2001.10853
Abstract
It is a challenging task to restore images from their variants with combined distortions. In the existing works, a promising strategy is to apply parallel "operations" to handle different types of distortion. However, in the feature fusion phase, a small number of operations would dominate the restoration result due to the features' heterogeneity by different operations. To this end, we introduce the tensor 1x1 convolutional layer by imposing high-order tensor (outer) product, by which we not only harmonize the heterogeneous features but also take additional non-linearity into account. To avoid the unacceptable kernel size resulted from the tensor product, we construct the kernels with tensor network decomposition, which is able to convert the exponential growth of the dimension to linear growth. Armed with the new layer, we propose High-order OWAN for multi-distorted image restoration. In the numerical experiments, the proposed net outperforms the previous state-of-the-art and shows promising performance even in more difficult tasks.
References in corpus (10)
- Adam: A Method for Stochastic Optimization
- Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
- SGDR: Stochastic Gradient Descent with Warm Restarts
- Tensor Ring Decomposition
- Tensor network decompositions in the presence of a global symmetry
- Path-Restore: Learning Network Path Selection for Image Restoration
- Efficient Low-rank Multimodal Fusion with Modality-Specific Factors
- Long-term Forecasting using Higher Order Tensor RNNs
- Single Image Super-resolution via a Lightweight Residual Convolutional Neural Network
- A^2Net: Adjacent Aggregation Networks for Image Raindrop Removal