Low-Rank Tensor Completion Based on Bivariate Equivalent Minimax-Concave Penalty

TL;DR

This paper introduces a novel bivariate equivalent minimax-concave penalty (BEMCP) for low-rank tensor completion, improving adaptability and performance in image and video data restoration tasks.

Contribution

It proposes the BEMCP theorem, derives its properties, and develops an ADMM-based method that outperforms existing approaches in practical applications.

Findings

Outperforms state-of-the-art methods in MSI, MRI, and CV data restoration

Provides new proximal operators for BEMCP and BEWTGN

Demonstrates improved adaptability to singular value changes

Abstract

Low-rank tensor completion (LRTC) is an important problem in computer vision and machine learning. The minimax-concave penalty (MCP) function as a non-convex relaxation has achieved good results in the LRTC problem. To makes all the constant parameters of the MCP function as variables so that futherly improving the adaptability to the change of singular values in the LRTC problem, we propose the bivariate equivalent minimax-concave penalty (BEMCP) theorem. Applying the BEMCP theorem to tensor singular values leads to the bivariate equivalent weighted tensor $\Gamma$-norm (BEWTGN) theorem, and we analyze and discuss its corresponding properties. Besides, to facilitate the solution of the LRTC problem, we give the proximal operators of the BEMCP theorem and BEWTGN. Meanwhile, we propose a BEMCP model for the LRTC problem, which is optimally solved based on alternating direction multiplier (ADMM). Finally, the proposed method is applied to the data restorations of multispectral image (MSI), magnetic resonance imaging (MRI) and color video (CV) in real-world, and the experimental results demonstrate that it outperforms the state-of-arts methods.