Open tkuri opened 3 years ago
顔のサーマル映像から人間のストレス状態を推定する。時空間のNNを利用しISTIという心臓の交感神経活動の変化を表す指標(ストレスと相関が高いと言われている)を再構成し、得られたISTIを用いてストレス状態を分類する。平均精度0.842でストレスを検出することが可能。
https://openaccess.thecvf.com/content/WACV2021/html/Kumar_StressNet_Detecting_Stress_in_Thermal_Videos_WACV_2021_paper.html
https://github.com/UCSB-VRL/StressNet-Detecting-stress-from-thermal-videos
論文概要
顔のサーマル映像から人間のストレス状態を推定する。時空間のNNを利用しISTIという心臓の交感神経活動の変化を表す指標(ストレスと相関が高いと言われている)を再構成し、得られたISTIを用いてストレス状態を分類する。平均精度0.842でストレスを検出することが可能。
https://openaccess.thecvf.com/content/WACV2021/html/Kumar_StressNet_Detecting_Stress_in_Thermal_Videos_WACV_2021_paper.html
Code
https://github.com/UCSB-VRL/StressNet-Detecting-stress-from-thermal-videos