Few-shot Human Activity Recognition on Noisy Wearable Sensor Data

Shizhuo Deng, Wen Hua, Botao Wang, Guoren Wang, Xiaofang Zhou

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

11 Citations (Scopus)

Abstract

Most existing wearable sensor-based human activity recognition (HAR) models are trained on substantial labeled data. It is difficult for HAR to learn new-class activities unseen during training from a few samples. Very few researches of few-shot learning (FSL) have been done in HAR to address the above problem, though FSL has been widely used in computer vision tasks. Besides, it is impractical to annotate sensor data with accurate activity labels in real-life applications. The noisy labels have great negative effects on FSL due to the limited samples. The weakly supervised few-shot learning in HAR is challenging, significant but rarely researched in existing literature. In this paper, we propose an end-to-end Weakly supervised Prototypical Networks (WPN) to learn more latent information from noisy data with multiple instance learning (MIL). In MIL, the noisy instances (subsequences of segmentation) have different labels from the bag’s (segmentation’s) label. The prototype is the center of the instances in WPN rather than less discriminative bags, which determines the bag-level classification accuracy. To get the most representative instance-level prototype, we propose two strategies to refine the prototype by selecting high-probability instances same as their bag’s label iteratively based on the distance-metric. The model is trained by minimizing the instance-level loss function and infers the final bag-level labels from instance-level labels. In the experiments, our proposals outperform existing approaches and achieve higher average ranks.

Original languageEnglish
Title of host publicationDatabase Systems for Advanced Applications - 25th International Conference, DASFAA 2020, Proceedings
EditorsYunmook Nah, Bin Cui, Sang-Won Lee, Jeffrey Xu Yu, Yang-Sae Moon, Steven Euijong Whang
PublisherSpringer Science and Business Media Deutschland GmbH
Pages54-72
Number of pages19
ISBN (Print)9783030594152
DOIs
Publication statusPublished - 2020
Externally publishedYes
Event25th International Conference on Database Systems for Advanced Applications, DASFAA 2020 - Jeju, Korea, Republic of
Duration: 24 Sept 202027 Sept 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12113 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference25th International Conference on Database Systems for Advanced Applications, DASFAA 2020
Country/TerritoryKorea, Republic of
CityJeju
Period24/09/2027/09/20

Keywords

  • Few-shot learning
  • Human activity recognition
  • Weakly supervised models
  • Wearable sensors

ASJC Scopus subject areas

  • Information Systems

Fingerprint

Dive into the research topics of 'Few-shot Human Activity Recognition on Noisy Wearable Sensor Data'. Together they form a unique fingerprint.

Cite this