Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.

As video data volumes soar exponentially, the significance of video content analysis, particularly Human Action Recognition (HAR), has become increasingly prominent in fields such as intelligent surveillance, sports analytics, medical rehabilitation, and virtual reality. However, current deep learni...

Full description

Saved in:
Bibliographic Details
Main Authors: Jun Qin, Shenwei Chen, Zheng Ye, Jing Liu, Zhou Liu
Format: Article
Language:English
Published: Public Library of Science (PLoS) 2025-01-01
Series:PLoS ONE
Online Access:https://doi.org/10.1371/journal.pone.0327717
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849427434591485952
author Jun Qin
Shenwei Chen
Zheng Ye
Jing Liu
Zhou Liu
author_facet Jun Qin
Shenwei Chen
Zheng Ye
Jing Liu
Zhou Liu
author_sort Jun Qin
collection DOAJ
description As video data volumes soar exponentially, the significance of video content analysis, particularly Human Action Recognition (HAR), has become increasingly prominent in fields such as intelligent surveillance, sports analytics, medical rehabilitation, and virtual reality. However, current deep learning-based HAR methods encounter challenges in recognizing subtle actions within complex backgrounds, comprehending long-term semantics, and maintaining computational efficiency. To address these challenges, we introduce the Video Swin-CLSTM Transformer. Based on the Video Swin Transformer backbone, our model incorporates optical flow information at the input stage to effectively counteract background interference, employing a sparse sampling strategy. Combined with the backbone's 3D Patch Partition and Patch Merging techniques, it efficiently extracts and fuses multi-level features from both optical flow and raw RGB inputs, thereby enhancing the model's ability to capture motion characteristics in complex backgrounds. Additionally, by embedding Convolutional Long Short-Term Memory (ConvLSTM) units, the model's capacity to capture and understand long-term dependencies among key actions in videos is further enhanced. Experiments on the UCF-101 dataset demonstrate that our model achieves mean Top-1/Top-5 accuracies of 92.8% and 99.4%, which are 3.2% and 2.0% higher than those of the baseline model, while the computational cost is reduced by an average of 3.3% at peak performance compared to models without optical flow. Ablation studies further validate the effectiveness of our model's crucial components, with the integration of optical flow and the embedding of ConvLSTM modules yielding maximum improvements in mean Top-1 accuracy of 2.6% and 1.9%, respectively. Notably, employing our custom ImageNet-1K-LSTM pre-training model results in a maximum increase of 2.7% in mean Top-1 accuracy compared to traditional ImageNet-1K pre-training model. These experimental results indicate that our model offers certain advantages over other Swin Transformer-based methods for video HAR tasks.
format Article
id doaj-art-9d5cbd9eea5844269985819eabefb783
institution Kabale University
issn 1932-6203
language English
publishDate 2025-01-01
publisher Public Library of Science (PLoS)
record_format Article
series PLoS ONE
spelling doaj-art-9d5cbd9eea5844269985819eabefb7832025-08-20T03:29:02ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01207e032771710.1371/journal.pone.0327717Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.Jun QinShenwei ChenZheng YeJing LiuZhou LiuAs video data volumes soar exponentially, the significance of video content analysis, particularly Human Action Recognition (HAR), has become increasingly prominent in fields such as intelligent surveillance, sports analytics, medical rehabilitation, and virtual reality. However, current deep learning-based HAR methods encounter challenges in recognizing subtle actions within complex backgrounds, comprehending long-term semantics, and maintaining computational efficiency. To address these challenges, we introduce the Video Swin-CLSTM Transformer. Based on the Video Swin Transformer backbone, our model incorporates optical flow information at the input stage to effectively counteract background interference, employing a sparse sampling strategy. Combined with the backbone's 3D Patch Partition and Patch Merging techniques, it efficiently extracts and fuses multi-level features from both optical flow and raw RGB inputs, thereby enhancing the model's ability to capture motion characteristics in complex backgrounds. Additionally, by embedding Convolutional Long Short-Term Memory (ConvLSTM) units, the model's capacity to capture and understand long-term dependencies among key actions in videos is further enhanced. Experiments on the UCF-101 dataset demonstrate that our model achieves mean Top-1/Top-5 accuracies of 92.8% and 99.4%, which are 3.2% and 2.0% higher than those of the baseline model, while the computational cost is reduced by an average of 3.3% at peak performance compared to models without optical flow. Ablation studies further validate the effectiveness of our model's crucial components, with the integration of optical flow and the embedding of ConvLSTM modules yielding maximum improvements in mean Top-1 accuracy of 2.6% and 1.9%, respectively. Notably, employing our custom ImageNet-1K-LSTM pre-training model results in a maximum increase of 2.7% in mean Top-1 accuracy compared to traditional ImageNet-1K pre-training model. These experimental results indicate that our model offers certain advantages over other Swin Transformer-based methods for video HAR tasks.https://doi.org/10.1371/journal.pone.0327717
spellingShingle Jun Qin
Shenwei Chen
Zheng Ye
Jing Liu
Zhou Liu
Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.
PLoS ONE
title Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.
title_full Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.
title_fullStr Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.
title_full_unstemmed Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.
title_short Video swin-CLSTM transformer: Enhancing human action recognition with optical flow and long-term dependencies.
title_sort video swin clstm transformer enhancing human action recognition with optical flow and long term dependencies
url https://doi.org/10.1371/journal.pone.0327717
work_keys_str_mv AT junqin videoswinclstmtransformerenhancinghumanactionrecognitionwithopticalflowandlongtermdependencies
AT shenweichen videoswinclstmtransformerenhancinghumanactionrecognitionwithopticalflowandlongtermdependencies
AT zhengye videoswinclstmtransformerenhancinghumanactionrecognitionwithopticalflowandlongtermdependencies
AT jingliu videoswinclstmtransformerenhancinghumanactionrecognitionwithopticalflowandlongtermdependencies
AT zhouliu videoswinclstmtransformerenhancinghumanactionrecognitionwithopticalflowandlongtermdependencies