The GIST and RIST of Iterative Self-Training for Semi-Supervised Segmentation

2021/03/31 by Eu Wern Teh, Terrance DeVries, Teh, Eu Wern +9 · 1 citation
Computer Science · #Domain Adaptation and Few-Shot Learning #Advanced Neural Network Applications #Advanced Image and Video Retrieval Techniques

paper · pdf · doi:10.48550/arxiv.2103.17105

Abstract

We consider the task of semi-supervised semantic segmentation, where we aim to produce pixel-wise semantic object masks given only a small number of human-labeled training examples. We focus on iterative self-training methods in which we explore the behavior of self-training over multiple refinement stages. We show that iterative self-training leads to performance degradation if done naïvely with a fixed ratio of human-labeled to pseudo-labeled training examples. We propose Greedy Iterative Self-Training (GIST) and Random Iterative Self-Training (RIST) strategies that alternate between training on either human-labeled data or pseudo-labeled data at each refinement stage, resulting in a performance boost rather than degradation. We further show that GIST and RIST can be combined with existing semi-supervised learning methods to boost performance.

Cited by

• FRIEREN: Federated Learning with Vision-Language Regularization for Segmentation

Related

• ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentation
• A Three-Stage Self-Training Framework for Semi-Supervised Semantic Segmentation
• Improving Semantic Segmentation via Self-Training
• Learning from Future: A Novel Self-Training Framework for Semantic Segmentation
• Improved Training for Self-Training by Confidence Assessments