Spatial-Temporal Transformer for Crime Recognition in Surveillance Videos

Kayleigh Boekhoudt; Estefania Talavera

doi:10.1109/AVSS56176.2022.9959414

Spatial-Temporal Transformer for Crime Recognition in Surveillance Videos

Kayleigh Boekhoudt
, Estefania Talavera

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

8 Citations (Scopus)

386 Downloads (Pure)

Abstract

Human-related crime recognition from surveillance videos becomes an even more challenging task when dealing with relatively similar human actions. We propose a transformer-based model that relies on the spatial-temporal representation of extracted skeletal trajectories for fine-grained classification. We validate the effectiveness of our model on the complex HR-Crime dataset consisting of videos representing 13 categories of human-related crimes. Quantitative and qualitative results suggest that building a transformer architecture with coupled spatial and temporal modules enables the model to compete in performance while improving intrinsic interpretability.

Original language	English
Title of host publication	AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance
Publisher	IEEE
Number of pages	8
ISBN (Electronic)	978-1-6654-6382-9
DOIs	https://doi.org/10.1109/AVSS56176.2022.9959414
Publication status	Published - 24 Nov 2022
Event	18th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2022 - Virtual, Online, Spain Duration: 29 Nov 2022 → 2 Dec 2022 Conference number: 18

Publication series

Name	AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance

Conference

Conference	18th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2022
Abbreviated title	AVSS 2022
Country/Territory	Spain
City	Virtual, Online
Period	29/11/22 → 2/12/22

Keywords

2023 OA procedure

Access to Document

10.1109/AVSS56176.2022.9959414

Spatial-Temporal_Transformer_for_Crime_Recognition_in_Surveillance_VideosFinal published version, 2.45 MBLicence: Taverne

Cite this

@inproceedings{6178383fe1d84c9f9bb59dea59d4a1f5,

title = "Spatial-Temporal Transformer for Crime Recognition in Surveillance Videos",

abstract = "Human-related crime recognition from surveillance videos becomes an even more challenging task when dealing with relatively similar human actions. We propose a transformer-based model that relies on the spatial-temporal representation of extracted skeletal trajectories for fine-grained classification. We validate the effectiveness of our model on the complex HR-Crime dataset consisting of videos representing 13 categories of human-related crimes. Quantitative and qualitative results suggest that building a transformer architecture with coupled spatial and temporal modules enables the model to compete in performance while improving intrinsic interpretability.",

keywords = "2023 OA procedure",

author = "Kayleigh Boekhoudt and Estefania Talavera",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2022, AVSS 2022 ; Conference date: 29-11-2022 Through 02-12-2022",

year = "2022",

month = nov,

day = "24",

doi = "10.1109/AVSS56176.2022.9959414",

language = "English",

series = "AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance",

publisher = "IEEE",

booktitle = "AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance",

address = "United States",

}

Boekhoudt, K & Talavera, E 2022, Spatial-Temporal Transformer for Crime Recognition in Surveillance Videos. in AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance. AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance, IEEE, 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2022, Virtual, Online, Spain, 29/11/22. https://doi.org/10.1109/AVSS56176.2022.9959414

Spatial-Temporal Transformer for Crime Recognition in Surveillance Videos. / Boekhoudt, Kayleigh; Talavera, Estefania.
AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance. IEEE, 2022. (AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Spatial-Temporal Transformer for Crime Recognition in Surveillance Videos

AU - Boekhoudt, Kayleigh

AU - Talavera, Estefania

N1 - Conference code: 18

PY - 2022/11/24

Y1 - 2022/11/24

N2 - Human-related crime recognition from surveillance videos becomes an even more challenging task when dealing with relatively similar human actions. We propose a transformer-based model that relies on the spatial-temporal representation of extracted skeletal trajectories for fine-grained classification. We validate the effectiveness of our model on the complex HR-Crime dataset consisting of videos representing 13 categories of human-related crimes. Quantitative and qualitative results suggest that building a transformer architecture with coupled spatial and temporal modules enables the model to compete in performance while improving intrinsic interpretability.

AB - Human-related crime recognition from surveillance videos becomes an even more challenging task when dealing with relatively similar human actions. We propose a transformer-based model that relies on the spatial-temporal representation of extracted skeletal trajectories for fine-grained classification. We validate the effectiveness of our model on the complex HR-Crime dataset consisting of videos representing 13 categories of human-related crimes. Quantitative and qualitative results suggest that building a transformer architecture with coupled spatial and temporal modules enables the model to compete in performance while improving intrinsic interpretability.

KW - 2023 OA procedure

UR - https://www.scopus.com/pages/publications/85143902115

U2 - 10.1109/AVSS56176.2022.9959414

DO - 10.1109/AVSS56176.2022.9959414

M3 - Conference contribution

AN - SCOPUS:85143902115

T3 - AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance

BT - AVSS 2022 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance

PB - IEEE

T2 - 18th IEEE International Conference on Advanced Video and Signal-Based Surveillance, AVSS 2022

Y2 - 29 November 2022 through 2 December 2022

ER -

Spatial-Temporal Transformer for Crime Recognition in Surveillance Videos

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this