Abstract:With the help of auxiliary data, tensor completion may better recover a low rank multidimensional array from limited observation entries. Most existing methods, including coupled matrix-tensor factorization and coupled tensor rank minimization, mainly focus on how to extract and incorporate subspace or directly use auxiliary data for tensor completion. They are either sensitive to a given rank or lack of physical interpretations of subspace information. In addition, the shared subspace information receives little attention in current tensor completion methods, especially there is no analysis of its impact on sample complexity. In this paper, we propose to separately explore and exploit shared subspaces for tensor completion. Specifically, dictionary learning takes the subspace from auxiliary data in the first step. Then a low rank optimization model for tensor completion is provided to incorporate the trained subspace by assuming that the recovered tensor is composed of two low rank components where one shares the subspace information with auxiliary data and the other is outside the shared space. Based on this optimization model, we make a quantitative analysis to illustrate the effect of subspace information on sample complexity, and provide theoretical insights into the usefulness of subspace information. Finally, experiments on simulated data are conducted to validate the theoretical analysis on the impact of subspace information. Experiments in two real-world applications including color image and multispectral image recovery show that the proposed method outperforms state-of-the-art ones in terms of prediction accuracy and CPU time.

Image tag completion via dual-view linear sparse reconstructions

Low-rank image tag completion with dual reconstruction structure preserved.

Image Tag Completion via Image-Specific and Tag-Specific Linear Sparse Reconstructions

Image Tag Completion by Low-Rank Factorization with Dual Reconstruction Structure Preserved

Rethinking Visual Reconstruction: Experience-Based Content Completion Guided by Visual Cues

Image Tag Completion and Refinement by Subspace Clustering and Matrix Completion

View-Guided Point Cloud Completion

Structure-Aware Image Completion with Texture Propagation

Image tag completion by local learning

Subspace Clustering Based Tag Sharing for Inductive Tag Matrix Refinement with Complex Errors

Tag Completion with Defective Tag Assignments Via Image-Tag Re-Weighting

Automatic and accurate image completion from a large displacement view

A Locality Sensitive Low-Rank Model for Image Tag Completion

Non-Linear Matrix Completion for Social Image Tagging.

Discovering Visual Concept Structure with Sparse and Incomplete Tags

A novel image tag completion method based on convolutional neural network

Semantic Reconstruction based on RGB Image and Sparse Depth

Visual Data Completion Via Local Sensitive Low Rank Tensor Learning.

Trainable Subspaces for Low Rank Tensor Completion: Model and Analysis

Context sensitive tag expansion with information inference

Point Cloud Scene Completion With Joint Color and Semantic Estimation From Single RGB-D Image