High-Quality Self-Supervised Deep Image Denoising
arXiv:1901.10277
Abstract
We describe a novel method for training high-quality image denoising models based on unorganized collections of corrupted images. The training does not need access to clean reference images, or explicit pairs of corrupted images, and can thus be applied in situations where such data is unacceptably expensive or impossible to acquire. We build on a recent technique that removes the need for reference data by employing networks with a "blind spot" in the receptive field, and significantly improve two key aspects: image quality and training efficiency. Our result quality is on par with state-of-the-art neural network denoisers in the case of i.i.d. additive Gaussian noise, and not far behind with Poisson and impulse noise. We also successfully handle cases where parameters of the noise model are variable and/or unknown in both training and evaluation data.
arXiv commentNeurIPS 2019 final version
References in corpus (13)
- Large Scale GAN Training for High Fidelity Natural Image Synthesis
- What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
- Conditional Image Generation with PixelCNN Decoders
- Pixel Recurrent Neural Networks
- Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
- Image Restoration Using Convolutional Auto-encoders with Symmetric Skip Connections
- There Are Many Consistent Explanations of Unlabeled Data: Why You Should Average
- Noise2Self: Blind Denoising by Self-Supervision
- Residual Dense Network for Image Restoration
- Unsupervised Learning with Stein's Unbiased Risk Estimator
- Multi-channel Weighted Nuclear Norm Minimization for Real Color Image Denoising
- Deep Learning with Inaccurate Training Data for Image Restoration
- An approach to image denoising using manifold approximation without clean images