國立虎尾科技大學 |

TAASR : = Temporally-Aware Affective State Recognition With Attention-Augmented CNNs.

Record Type:	Language materials, manuscript : Monograph/item
Title/Author:	TAASR :/
Reminder of title:	Temporally-Aware Affective State Recognition With Attention-Augmented CNNs.
Author:	Abramow, Daniel.
Description:	1 online resource (48 pages)
Notes:	Source: Masters Abstracts International, Volume: 84-12.
Contained By:	Masters Abstracts International84-12.
Subject:	Mental health. -
Online resource:	click for full text (PQDT)
ISBN:	9798379687311

TAASR : = Temporally-Aware Affective State Recognition With Attention-Augmented CNNs.
Abramow, Daniel.

TAASR :Temporally-Aware Affective State Recognition With Attention-Augmented CNNs. - 1 online resource (48 pages)

Source: Masters Abstracts International, Volume: 84-12.

Thesis (M.S.)--University of Georgia, 2023.

Includes bibliographical references

Affective state recognition (ASR) involves using the body's physiological signals to extract useful information about ones mental state. ASR systems are often implemented in highly controlled environments with cumbersome chest sensors and intrusive facial expression monitoring setups, making it difficult to translate their performance to real environments. Recently, the widespread adoption of wrist-worn wearables has highlighted a need for further research into practical ASR with commercially available devices. In this paper, we propose TAASR, an InceptionTime based end-to-end learning architecture augmented with channel attention and global feature fusion for three-class ASR (baseline vs. stressed vs. amused), and TAASR-MT, a multi-task version of TAASR that uses mental health self-assessments to improve basic ASR performance. For practicality, we train these architectures primarily with wrist-based signals and report a best classification accuracy and F1-score of 81.16% and 70.02, demonstrating noticeable improvements upon InceptionTime and prior works that employ simpler classification approaches.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798379687311Subjects--Topical Terms:

564038
Mental health.
Subjects--Index Terms:

Affective state recognitionIndex Terms--Genre/Form:

554714
Electronic books.

TAASR : = Temporally-Aware Affective State Recognition With Attention-Augmented CNNs.
LDR:02524ntm a22004097 4500 001 1149293
005 20241015112503.5
006 m o d
007 cr bn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798379687311
035 $a (MiAaPQ)AAI30311305
035 $a AAI30311305
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Abramow, Daniel. $3 1475465
245 1 0 $a TAASR : $b Temporally-Aware Affective State Recognition With Attention-Augmented CNNs.
264 0 $c 2023
300 $a 1 online resource (48 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 84-12.
500 $a Advisor: Liu, Tianming.
502 $a Thesis (M.S.)--University of Georgia, 2023.
504 $a Includes bibliographical references
520 $a Affective state recognition (ASR) involves using the body's physiological signals to extract useful information about ones mental state. ASR systems are often implemented in highly controlled environments with cumbersome chest sensors and intrusive facial expression monitoring setups, making it difficult to translate their performance to real environments. Recently, the widespread adoption of wrist-worn wearables has highlighted a need for further research into practical ASR with commercially available devices. In this paper, we propose TAASR, an InceptionTime based end-to-end learning architecture augmented with channel attention and global feature fusion for three-class ASR (baseline vs. stressed vs. amused), and TAASR-MT, a multi-task version of TAASR that uses mental health self-assessments to improve basic ASR performance. For practicality, we train these architectures primarily with wrist-based signals and report a best classification accuracy and F1-score of 81.16% and 70.02, demonstrating noticeable improvements upon InceptionTime and prior works that employ simpler classification approaches.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Mental health. $3 564038
650 4 $a Information science. $3 561178
653 $a Affective state recognition
653 $a Convolutional neural networks
653 $a Deep learning
653 $a Time series classification
653 $a Wearables
653 $a InceptionTime
655 7 $a Electronic books. $2 local $3 554714
690 $a 0800
690 $a 0723
690 $a 0347
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a University of Georgia. $b Artificial Intelligence - MS. $3 1466144
773 0 $t Masters Abstracts International $g 84-12.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30311305 $z click for full text (PQDT)