Tensor Methods in Computer Vision and Deep Learning

Abstract

Tensors, or multidimensional arrays, are data structures that can naturally represent visual data of multiple dimensions. Inherently able to efficiently capture structured, latent semantic spaces and high-order interactions, tensors have a long history of applications in a wide span of computer vision problems. With the advent of the deep learning paradigm shift in computer vision, tensors have become even more fundamental. Indeed, essential ingredients in modern deep learning architectures, such as convolutions and attention mechanisms, can readily be considered as tensor mappings. In effect, tensor methods are increasingly finding significant applications in deep learning, including the design of memory and compute efficient network architectures, improving robustness to random noise and adversarial attacks, and aiding the theoretical understanding of deep networks. This article provides an in-depth and practical review of tensors and tensor methods in the context of representation learning and deep learning, with a particular focus on visual data analysis and computer vision applications. Concretely, besides fundamental work in tensor-based visual data analysis methods, we focus on recent developments that have brought on a gradual increase of tensor methods, especially in deep learning architectures, and their implications in computer vision applications. To further enable the newcomer to grasp such concepts quickly, we provide companion Python notebooks, covering key aspects of the paper and implementing them, step-by-step with TensorLy.

arXiv commentProceedings of the IEEE (2021)

References in corpus (48)

• Auto-Encoding Variational Bayes
• Neural Machine Translation by Jointly Learning to Align and Translate
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems
• Tensor Decomposition for Signal Processing and Machine Learning
• Tensor Decompositions for Signal Processing Applications From Two-way to Multiway Component Analysis
• A survey of sparse representation: algorithms and applications
• The ITensor Software Library for Tensor Network Calculations
• Bayesian CP Factorization of Incomplete Tensors with Automatic Rank Determination
• Tensor Regression with Applications in Neuroimaging Data Analysis
• Highway Networks
• A class of ansatz wave functions for 1D spin systems and their relation to DMRG
• Robust Low-rank Tensor Recovery: Models and Algorithms
• Tensor Ring Decomposition
• Tensor Networks for Dimensionality Reduction and Large-Scale Optimizations. Part 2 Applications and Future Perspectives
• TensorLy: Tensor Learning in Python
• Linked Component Analysis from Matrices to High Order Tensors: Applications to Biomedical Data
• TTHRESH: Tensor Compression for Multidimensional Visual Data
• Beating the Perils of Non-Convexity: Guaranteed Training of Neural Networks using Tensor Methods
• Tensor-Based Classifiers for Hyperspectral Data Analysis
• Learning Functions: When Is Deep Better Than Shallow
• Deep Multi-task Representation Learning: A Tensor Factorisation Approach
• Ultimate tensorization: compressing convolutional and FC layers alike
• Global Optimality in Tensor Factorization, Deep Learning, and Beyond
• Convolutional Tensor-Train LSTM for Spatio-temporal Learning
• Deep Polynomial Neural Networks
• Tensor Regression Networks
• Multi-Head Attention: Collaborate Instead of Concatenate
• Decoupling Multivariate Polynomials Using First-Order Information
• Tensor Contractions with Extended BLAS Kernels on CPU and GPU
• Higher-Order Factorization Machines
• Dictionary Learning for Massive Matrix Factorization
• Tensor Train decomposition on TensorFlow (T3F)
• Spectral Learning on Matrices and Tensors
• Learning from Multiway Data: Simple and Efficient Tensor Regression
• Minimax Lower Bounds on Dictionary Learning for Tensor Data
• An Adversarial Neuro-Tensorial Approach For Learning Disentangled Representations
• Learning Efficient Tensor Representations with Ring Structure Networks
• Tensor Regression Networks with various Low-Rank Tensor Approximations
• A New Low-Rank Tensor Model for Video Completion
• Minimax Lower Bounds for Kronecker-Structured Dictionary Learning
• Compositional Hierarchical Tensor Factorization: Representing Hierarchical Intrinsic and Extrinsic Causal Factors
• tntorch: Tensor Network Learning with PyTorch
• Understanding Generalization in Deep Learning via Tensor Methods
• Tensor Networks for Medical Image Classification
• Tensor Contraction Layers for Parsimonious Deep Nets
• Multilinear Latent Conditioning for Generating Unseen Attribute Combinations
• TensorShield: Tensor-based Defense Against Adversarial Attacks on Images

Cited by in corpus (14)

• Guaranteed Tensor Recovery Fused Low-rankness and Smoothness
• Rethinking Bayesian Learning for Data Analysis: The Art of Prior and Inference in Sparsity-Aware Modeling
• Tensor-based Basis Function Learning for Three-dimensional Sound Speed Fields
• A Comprehensive Review of Visual-Textual Sentiment Analysis from Social Media Networks
• Context in object detection: a systematic literature review
• Reduced-Rank Tensor-on-Tensor Regression and Tensor-variate Analysis of Variance
• TedNet: A Pytorch Toolkit for Tensor Decomposition Networks
• Variational Tensor Neural Networks for Deep Learning
• Positive unlabeled learning with tensor networks
• The Tensor Track VII: From Quantum Gravity to Artificial Intelligence
• Reinforcement Learning in Factored Action Spaces using Tensor Decompositions
• Ricci-Notation Tensor Framework for Model-based Approaches to Imaging
• tCURLoRA: Tensor CUR Decomposition Based Low-Rank Parameter Adaptation and Its Application in Medical Image Segmentation
• Multi-Scale Tensorial Summation and Dimensional Reduction Guided Neural Network for Edge Detection