Applying deep incremental learning-based posture recognition model for injury prevention in construction

Junqi Zhao, Esther Obonyo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This research investigated the feasibility of applying a Deep Incremental Learning model to achieve a high posture recognition performance with Wearable Sensors (WS). The authors use the recognition of Musculoskeletal Disorder (MSDs) related postures among construction workers as the testbed. This research proposed the Convolutional Long Short-Term Memory (CLN) model under Incremental Learning (IL), where a trained model adapts to new subject' postures to maintain high recognition performance. The model was evaluated on datasets from nine construction workers. Results show: i) CLN model with shallow convolutional layers achieved high recognition accuracy (Macro F1 score) under personalized (0.87) and generalized (0.84) modelling; ii) Generalized CLN model under “Many-to-One” IL strategy can adapt to a new subject and balance the forgetting of learnt subjects; iii) incremental CLN model gave close detection of posture proportion and holding time to ground-truth, which facilitates reliable MSDs risk assessment and further prevention through monitoring injury-related postures.

Original languageEnglish (US)
Title of host publicationEG-ICE 2020 Workshop on Intelligent Computing in Engineering, Proceedings
EditorsLucian-Constantin Ungureanu, Timo Hartmann
PublisherUniversitatsverlag der TU Berlin
Pages93-105
Number of pages13
ISBN (Electronic)9783798331556
StatePublished - 2020
Event27th EG-ICE International Workshop on Intelligent Computing in Engineering 2020 - Virtual, Online, Germany
Duration: Jul 1 2020Jul 4 2020

Publication series

NameEG-ICE 2020 Workshop on Intelligent Computing in Engineering, Proceedings

Conference

Conference27th EG-ICE International Workshop on Intelligent Computing in Engineering 2020
CountryGermany
CityVirtual, Online
Period7/1/207/4/20

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Engineering(all)

Fingerprint Dive into the research topics of 'Applying deep incremental learning-based posture recognition model for injury prevention in construction'. Together they form a unique fingerprint.

Cite this