Domain generalization for activity recognition: Learn from visible, infer with thermal - Archive ouverte HAL Access content directly
Conference Papers Year : 2022

Domain generalization for activity recognition: Learn from visible, infer with thermal

Abstract

We proposed a solution based on I3D and optical flow to learn common characteristics between thermal and visible videos. For this purpose we proposed a new database to evaluate our solution. The new model comprises an optical flow extractor; a feature extractor based on I3D, a domain classifier, and an activity recognition classifier. We learn invariant characteristics computed from the optical flow. We have simulated several source domains, and we have shown that it is possible to obtain excellent results on a modality that was not used during the training. Such techniques can be used when there is only one source and one target domain.
Fichier principal
Vignette du fichier
ICPRAM2022.pdf (707.47 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03588563 , version 1 (25-02-2022)

Identifiers

Cite

Yannick Zoetgnande, Jean-Louis Dillenseger. Domain generalization for activity recognition: Learn from visible, infer with thermal. 11th International Conference on Pattern Recognition Applications and Methods, Feb 2022, Online Streaming, France. pp.722-729, ⟨10.5220/0010906300003122⟩. ⟨hal-03588563⟩
93 View
169 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More