Non-Stationary Predictions May Be More Informative: Exploring Pseudo-Labels with a Two-Phase Pattern of Training Dynamics

Hongbin Pei; Jingxin Hai; Yu Li; Huiqi Deng; Denghao Ma; Jie Ma; Pinghui Wang; Jing Tao; Xiaohong Guan

2025 ICML ICML 2025

Non-Stationary Predictions May Be More Informative: Exploring Pseudo-Labels with a Two-Phase Pattern of Training Dynamics

Abstract

No abstract available. No abstract was captured for this paper when its proceedings were scraped - most often an older edition published without one online. How the corpus is built.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Computer Vision

Authors

Hongbin Pei, Jingxin Hai, Yu Li, Huiqi Deng, Denghao Ma, Jie Ma, Pinghui Wang, Jing Tao, Xiaohong Guan

Topics

Machine Learning > Core Methods > Classification Computer Vision > Analysis > Object Detection Deep Learning > Learning Types > Self-Supervised Learning

Keywords

spatial pattern loss function temporal pattern pseudo-labelings 2-phasic metric two-phase labels

Download PDF