Bo Song

BadLabel: A Robust Perspective on Evaluating and Enhancing Label-Noise Learning

2024 · IEEE Transactions on Pattern Analysis and Machine Intelligence

Label-noise learning (LNL) aims to increase the model's generalization given training data with noisy labels. To facilitate practical LNL algorithms, researchers have proposed different label noise types, ranging from class-conditional to instance-dependent noises. In this …

Reinforcement Learning with Decoupled State Representation for Robot Manipulations

2024 · Neural Processing Letters

Abstract Deep reinforcement learning has significantly advanced robot manipulations by providing an alternative solution for designing control strategies using raw images as direct inputs. While images offer additional environmental information, the end-to-end policy training manner …