papers

Publications (8)

cs.LG2025

PTMs-TSCIL Pre-Trained Models Based Class-Incremental Learning

Yuanlong Wu, Mingxing Nie, Tao Zhu +3

Class-incremental learning (CIL) for time series data faces critical challenges in balancing stability against catastrophic forgetting and plasticity for new knowledge acquisition,…

cs.LG2024

MCformer: Multivariate Time Series Forecasting with Mixed-Channels Transformer

Wenyong Han, Tao Zhu Member, Liming Chen +3

The massive generation of time-series data by largescale Internet of Things (IoT) devices necessitates the exploration of more effective models for multivariate time-series forecas…

cs.HC2022

Sensor Data Augmentation by Resampling for Contrastive Learning in Human Activity Recognition

Jinqiang Wang, Tao Zhu, Jingyuan Gan +3

While deep learning has contributed to the advancement of sensor-based Human Activity Recognition (HAR), it is usually a costly and challenging supervised task with the needs of a…

cs.CV2024

HARMamba: Efficient and Lightweight Wearable Sensor Human Activity Recognition Based on Bidirectional Mamba

Shuangjian Li, Tao Zhu, Furong Duan +4

Wearable sensor-based human activity recognition (HAR) is a critical research domain in activity perception. However, achieving high efficiency and long sequence recognition remain…

cs.CV2022

Negative Selection by Clustering for Contrastive Learning in Human Activity Recognition

Jinqiang Wang, Tao Zhu, Liming Chen +2

Contrastive learning has been applied to Human Activity Recognition (HAR) based on sensor data owing to its ability to achieve performance comparable to supervised learning with a…

cs.LG2025

HAR-DoReMi: Optimizing Data Mixture for Self-Supervised Human Activity Recognition Across Heterogeneous IMU Datasets

Lulu Ban, Tao Zhu, Xiangqing Lu +7

Cross-dataset Human Activity Recognition (HAR) suffers from limited model generalization, hindering its practical deployment. To address this critical challenge, inspired by the su…