Počet záznamů: 1
Interpretable Gait Recognition by Granger Causality
- 1.
SYSNO ASEP 0567047 Druh ASEP C - Konferenční příspěvek (mezinárodní konf.) Zařazení RIV D - Článek ve sborníku Název Interpretable Gait Recognition by Granger Causality Tvůrce(i) Balazia, M. (FR)
Hlaváčková-Schindler, Kateřina (UIVT-O) RID, ORCID, SAI
Sojka, P. (CZ)
Plant, C. (AT)Celkový počet autorů 4 Zdroj.dok. 2022 26th International Conference on Pattern Recognition (ICPR). - Piscataway : IEEE, 2022 - ISSN 1051-4651 - ISBN 978-1-6654-9063-4 Rozsah stran s. 1069-1075 Poč.str. 7 s. Forma vydání Tištěná - P Akce ICPR 2022: International Conference on Pattern Recognition /26./ Datum konání 21.08.2022 - 25.08.2022 Místo konání Montréal Země CA - Kanada Typ akce WRD Jazyk dok. eng - angličtina Země vyd. US - Spojené státy americké Klíč. slova Measurement ; Analytical models ; Three-dimensional displays ; Neural networks ; Video surveillance ; Skeleton ; Motion capture Obor OECD Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8) CEP GA19-16066S GA ČR - Grantová agentura ČR Institucionální podpora UIVT-O - RVO:67985807 UT WOS 000897707601011 EID SCOPUS 85143611376 DOI 10.1109/ICPR56361.2022.9956624 Anotace Which joint interactions in the human gait cycle can be used as biometric characteristics? Most current methods on gait recognition suffer from the lack of interpretability. We propose an interpretable feature representation of gait sequences by the graphical Granger causal inference. Gait sequence of a person in the standardized motion capture format, constituting a set of 3D joint spatial trajectories, is envisaged as a causal system of joints interacting in time. We apply the graphical Granger model (GGM) to obtain the so-called Granger causal graph among joints as a discriminative and visually interpretable representation of a person's gait. We evaluate eleven distance functions in the GGM feature space by established classification and class-separability evaluation metrics. Our experiments indicate that, depending on the metric, the most appropriate distance functions for the GGM are the total norm distance and the Ky-Fan 1-norm distance. Experiments also show that the GGM is able to detect the most discriminative joint interactions and that it outperforms five related interpretable models in correct classification rate and in Davies-Bouldin index. The proposed GGM model can serve as a complementary tool for gait analysis in kinesiology or for gait recognition in video surveillance. Pracoviště Ústav informatiky Kontakt Tereza Šírová, sirova@cs.cas.cz, Tel.: 266 053 800 Rok sběru 2023
Počet záznamů: 1