Self-Validation: Early Stopping for Single-Instance Deep Generative Priors
arXiv:2110.12271
Abstract
Recent works have shown the surprising effectiveness of deep generative models in solving numerous image reconstruction (IR) tasks, even without training data. We call these models, such as deep image prior and deep decoder, collectively as single-instance deep generative priors (SIDGPs). The successes, however, often hinge on appropriate early stopping (ES), which by far has largely been handled in an ad-hoc manner. In this paper, we propose the first principled method for ES when applying SIDGPs to IR, taking advantage of the typical bell trend of the reconstruction quality. In particular, our method is based on collaborative training and self-validation: the primal reconstruction process is monitored by a deep autoencoder, which is trained online with the historic reconstructed images and used to validate the reconstruction quality constantly. Experimentally, on several IR problems and different SIDGPs, our self-validation method is able to reliably detect near-peak performance and signal good ES points. Our code is available at https://sun-umn.github.io/Self-Validation/.
arXiv commentTo appear in British Machine Vision Conference (BMVC) 2021
References in corpus (7)
- Compressed Sensing using Generative Models
- Algorithmic Guarantees for Inverse Imaging with Untrained Network Priors
- Denoising and Regularization via Exploiting the Structural Bias of Convolutional Generators
- Solving Inverse Problems with Hybrid Deep Image Priors: the challenge of preventing overfitting
- Inverse Problems, Deep Learning, and Symmetry Breaking
- Phase Retrieval with Holography and Untrained Priors: Tackling the Challenges of Low-Photon Nanoscale Imaging
- Rethinking Deep Image Prior for Denoising