Self-supervised regression learning using domain knowledge: Applications to improving self-supervised denoising in imaging
This work addresses a gap in self-supervised learning for regression in computational imaging and computer vision, offering a domain-specific solution that is incremental in extending self-supervised techniques.
The paper tackles the lack of self-supervised learning methods for regression tasks beyond image denoising by proposing a general framework that uses domain knowledge via a pseudo-predictor, and it demonstrates significant improvements in denoising quality for low-dose CT and camera images over existing methods.
Regression that predicts continuous quantity is a central part of applications using computational imaging and computer vision technologies. Yet, studying and understanding self-supervised learning for regression tasks - except for a particular regression task, image denoising - have lagged behind. This paper proposes a general self-supervised regression learning (SSRL) framework that enables learning regression neural networks with only input data (but without ground-truth target data), by using a designable pseudo-predictor that encapsulates domain knowledge of a specific application. The paper underlines the importance of using domain knowledge by showing that under different settings, the better pseudo-predictor can lead properties of SSRL closer to those of ordinary supervised learning. Numerical experiments for low-dose computational tomography denoising and camera image denoising demonstrate that proposed SSRL significantly improves the denoising quality over several existing self-supervised denoising methods.