Deep neural networks based purely on attention have been successful across several domains, relying on minimal architectural priors from the designer. In Human Action Recognition (HAR), attention mechanisms have been primarily adopted on top of standard convolutional or recurrent layers, improving the overall generalization capability. In this work, we introduce Action Transformer (AcT), a simple, fully, self-attentional architecture that consistently outperforms more elaborated networks that mix convolutional, recurrent, and attentive layers. In order to limit computational and energy requests, building on previous human action recognition research, the proposed approach exploits 2D pose representations over small temporal windows, providing a low latency solution for accurate and effective real-time performance. Moreover, we open-source MPOSE2021, a new large-scale dataset, as an attempt to build a formal training and evaluation benchmark for real-time, short-time HAR. The proposed methodology was extensively tested on MPOSE2021 and compared to several state-of-the-art architectures, proving the effectiveness of the AcT model and laying the foundations for future work on HAR.

Action Transformer: A Self-Attention Model for Short-Time Pose-Based Human Action Recognition / Mazzia, Vittorio; Angarano, Simone; Salvetti, Francesco; Angelini, Federico; Chiaberge, Marcello. - In: PATTERN RECOGNITION. - ISSN 0031-3203. - ELETTRONICO. - 124:(2022), p. 108487. [10.1016/j.patcog.2021.108487]

Action Transformer: A Self-Attention Model for Short-Time Pose-Based Human Action Recognition

Mazzia, Vittorio;Angarano, Simone;Salvetti, Francesco;Chiaberge, Marcello
2022

Abstract

Deep neural networks based purely on attention have been successful across several domains, relying on minimal architectural priors from the designer. In Human Action Recognition (HAR), attention mechanisms have been primarily adopted on top of standard convolutional or recurrent layers, improving the overall generalization capability. In this work, we introduce Action Transformer (AcT), a simple, fully, self-attentional architecture that consistently outperforms more elaborated networks that mix convolutional, recurrent, and attentive layers. In order to limit computational and energy requests, building on previous human action recognition research, the proposed approach exploits 2D pose representations over small temporal windows, providing a low latency solution for accurate and effective real-time performance. Moreover, we open-source MPOSE2021, a new large-scale dataset, as an attempt to build a formal training and evaluation benchmark for real-time, short-time HAR. The proposed methodology was extensively tested on MPOSE2021 and compared to several state-of-the-art architectures, proving the effectiveness of the AcT model and laying the foundations for future work on HAR.
File in questo prodotto:
File Dimensione Formato  
AcT___Pattern_Recognition.pdf

Open Access dal 16/12/2023

Descrizione: Post-print
Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: Creative commons
Dimensione 7.19 MB
Formato Adobe PDF
7.19 MB Adobe PDF Visualizza/Apri
1-s2.0-S0031320321006634-main.pdf

non disponibili

Descrizione: Post-print versione editoriale
Tipologia: 2a Post-print versione editoriale / Version of Record
Licenza: Non Pubblico - Accesso privato/ristretto
Dimensione 2.88 MB
Formato Adobe PDF
2.88 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2946512