TY - GEN
T1 - Foreground Detection Using an Attention Module and a Video Encoding
AU - Benavides-Arce, Anthony A.
AU - Flores-Benites, Victor
AU - Mora-Colque, Rensso
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - Foreground detection is the task of labelling the foreground or background pixels in the video sequence and it depends on the context of the scene. For many years, methods based on background model have been the most used approaches for detecting foreground; however, their methods are sensitive to error propagation from the first background model estimations. To address this problem, we proposed a U-net based architecture with an attention module, where the encoding of the entire video sequence is used as attention context to get features related to the background model. We tested our network on sixteen scenes from the CDnet2014 dataset, with an average F-measure of 88.42. The results also show that our model outperforms traditional and neural networks methods. Thus, we demonstrated that an attention module on a U-net based architecture can deal with the foreground detection challenges.
AB - Foreground detection is the task of labelling the foreground or background pixels in the video sequence and it depends on the context of the scene. For many years, methods based on background model have been the most used approaches for detecting foreground; however, their methods are sensitive to error propagation from the first background model estimations. To address this problem, we proposed a U-net based architecture with an attention module, where the encoding of the entire video sequence is used as attention context to get features related to the background model. We tested our network on sixteen scenes from the CDnet2014 dataset, with an average F-measure of 88.42. The results also show that our model outperforms traditional and neural networks methods. Thus, we demonstrated that an attention module on a U-net based architecture can deal with the foreground detection challenges.
KW - Attention
KW - Foreground Detection
KW - U-Net
KW - Video encoding
UR - https://www.scopus.com/pages/publications/85131149770
U2 - 10.1007/978-3-031-06433-3_17
DO - 10.1007/978-3-031-06433-3_17
M3 - Conference contribution
AN - SCOPUS:85131149770
SN - 9783031064326
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 195
EP - 205
BT - Image Analysis and Processing – ICIAP 2022 - 21st International Conference, 2022, Proceedings
A2 - Sclaroff, Stan
A2 - Distante, Cosimo
A2 - Leo, Marco
A2 - Farinella, Giovanni M.
A2 - Tombari, Federico
PB - Springer Science and Business Media Deutschland GmbH
T2 - 21st International Conference on Image Analysis and Processing, ICIAP 2022
Y2 - 23 May 2022 through 27 May 2022
ER -