A Hybrid 3D CNNs Transformer Architecture for Video-Based Human Action Recognition with Improved Accuracy on UCF101 and HMDB51

Basit Öğe Kaydı
dc.contributor.authorSeven, Engin
dc.contributor.authorDemirel, Eylem Yücel
dc.date.accessioned2026-02-08T15:04:49Z
dc.date.available2026-02-08T15:04:49Z
dc.date.issued2025
dc.departmentBursa Teknik Üniversitesi
dc.description.abstractVideo-Based Human Action Recognition (HAR) remains challenging due to inter-class similarity, background noise, and the need to capture long-term temporal dependencies. This study proposes a hybrid deep learning model that integrates 3D Convolutional Neural Networks (3D CNNs) with Transformer-based attention mechanisms to jointly capture spatio-temporal features and long-range motion context. The architecture was optimized for parameter efficiency and trained on the UCF101 and HMDB51 benchmark datasets using standardized preprocessing and training strategies. Experimental results indicate that the proposed model reaches 97% accuracy and 96.8% mean F1-score on UCF101, and 85% accuracy, and 83.8% F1-score on HMDB51, showing consistent improvements compared to the standalone 3D CNNs and Transformer variants under identical settings. Ablation studies confirm that the combination of convolutional and attention layers significantly improves recognition performance while maintaining competitive computational cost (3.78M parameters, 17.75 GFLOPs/video, ~7 ms GPU latency). These findings highlight the effectiveness of the hybrid design for accurate and efficient HAR. Future work will address class imbalance using focal loss or weighted training, explore multimodal data integration, and develop more lightweight Transformer modules for real-time deployment on resource-constrained devices.
dc.identifier.doi10.38088/jise.1703936
dc.identifier.endpage342
dc.identifier.issn2602-4217
dc.identifier.issue2
dc.identifier.startpage327
dc.identifier.urihttps://doi.org/10.38088/jise.1703936
dc.identifier.urihttps://hdl.handle.net/20.500.12885/4224
dc.identifier.volume9
dc.language.isoen
dc.publisherBursa Teknik Üniversitesi
dc.relation.ispartofJournal of Innovative Science and Engineering
dc.relation.publicationcategoryMakale - Ulusal Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/openAccess
dc.snmzKA_DergiPark_20260207
dc.subjectImage Processing
dc.subjectGörüntü İşleme
dc.titleA Hybrid 3D CNNs Transformer Architecture for Video-Based Human Action Recognition with Improved Accuracy on UCF101 and HMDB51
dc.typeArticle

Dosyalar

Koleksiyon

Journal of Innovative Science and Engineering Koleksiyonu