Sparse Prior

Sparse priors are increasingly used in machine learning to improve model efficiency and performance, particularly in scenarios with limited data or high computational costs. Current research focuses on integrating sparse priors into various architectures, including deep learning models and vision transformers, often employing techniques like tensor decomposition and data-driven sparsification to enhance feature extraction and reduce computational burden. This approach shows promise in diverse applications, from improving medical image segmentation and infrared target detection to enabling real-time depth estimation for underwater robotics, demonstrating the broad applicability and impact of sparse prior methods. The effectiveness of sparse priors is being rigorously evaluated and theoretically grounded, with a growing focus on understanding their behavior in different model types and data contexts.

Papers

July 22, 2024

Sparse Prior Is Not All You Need: When Differential Directionality Meets Saliency Coherence for Infrared Small Target Detection
Fei Zhou, Maixia Fu, Yulei Qian, Jian Yang, Yimian Dai
Infrared Small Target Detection Target Detection Small Target Detection Information Theoretic Detectability Unaddressed Directionality Image Coherence Sparse Prior

July 5, 2024

Few-Shot Airway-Tree Modeling using Data-Driven Sparse Priors
Ali Keshavarzi, Elsa Angelini
Segmentation Model Sparse Representation Random Sparsification Sparse Learning Segmentation Pipeline Sparse Prior

May 22, 2024

Vision Transformer with Sparse Scan Prior
Qihang Fan, Huaibo Huang, Mingrui Chen, Ran He
Vision Transformer Semantic Segmentation Vision Model Computer Vision Task Feature Redundancy Sparse Prior

March 21, 2024

Posterior concentrations of fully-connected Bayesian neural networks with general priors on the weights
Insung Kong, Yongdai Kim
Bayesian Neural Network Balancing Weight SAM Prior Gaussian Prior Posterior Concentration Sparse Prior

October 25, 2023

Metrically Scaled Monocular Depth Estimation through Sparse Priors for Underwater Robots
Luca Ebner, Gideon Billings, Stefan Williams
Depth Estimation Depth Prediction Underwater Robot Scale Aware Depth Sparse Prior

Sparse Prior

Papers

Sparse Prior Is Not All You Need: When Differential Directionality Meets Saliency Coherence for Infrared Small Target Detection

Few-Shot Airway-Tree Modeling using Data-Driven Sparse Priors

Vision Transformer with Sparse Scan Prior

Posterior concentrations of fully-connected Bayesian neural networks with general priors on the weights

Metrically Scaled Monocular Depth Estimation through Sparse Priors for Underwater Robots