Semi-Supervised Semantic Segmentation Based on Pseudo-Labels: A Survey
arXiv:2403.01909 · doi:10.1109/TCSVT.2024.3508768
Abstract
Semantic segmentation is an important and popular research area in computer vision that focuses on classifying pixels in an image based on their semantics. However, supervised deep learning requires large amounts of data to train models and the process of labeling images pixel by pixel is time-consuming and laborious. This review aims to provide a first comprehensive and organized overview of the state-of-the-art research results on pseudo-label methods in the field of semi-supervised semantic segmentation, which we categorize from different perspectives and present specific methods for specific application areas. In addition, we explore the application of pseudo-label technology in medical and remote-sensing image segmentation. Finally, we also propose some feasible future research directions to address the existing challenges.
Accepted by IEEE Transactions on Circuits and Systems for Video Technology(TCSVT)
References in corpus (17)
- Learning to Prompt for Vision-Language Models
- The Medical Segmentation Decathlon
- The Liver Tumor Segmentation Benchmark (LiTS)
- Skin Lesion Analysis Toward Melanoma Detection 2018: A Challenge Hosted by the International Skin Imaging Collaboration (ISIC)
- Multi-Task Learning for Dense Prediction Tasks: A Survey
- Uncertainty-aware multi-view co-training for semi-supervised medical image segmentation and domain adaptation
- DMT: Dynamic Mutual Training for Semi-Supervised Learning
- Semantic Understanding of Foggy Scenes with Purely Synthetic Data
- Compete to Win: Enhancing Pseudo Labels for Barely-supervised Medical Image Segmentation
- Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation
- Learning from Pixel-Level Label Noise: A New Perspective for Semi-Supervised Semantic Segmentation
- A Survey on Semi-Supervised Semantic Segmentation
- Semi-Supervised Semantic Segmentation with Cross Teacher Training
- Semi-Supervised Building Footprint Generation with Feature and Output Consistency Training
- DDF: A Novel Dual-Domain Image Fusion Strategy for Remote Sensing Image Semantic Segmentation with Unsupervised Domain Adaptation
- Revisiting consistency for semi-supervised semantic segmentation
- Picking Up Quantization Steps for Compressed Image Classification