KCC 2021
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
´ë±Ô¸ð ¿µ»ó¿¡¼ ¹ßÇöµÈ Ç¥ÇöÀνÄÀ» À§ÇÑ ¸ÖƼ¸ð´Þ µö ±×·¡ÇÁ |
¿µ¹®Á¦¸ñ(English Title) |
Multimodal Deep Graph Fusion for Evoked Expression Recognition in Large-Scale Videos |
ÀúÀÚ(Author) |
Æ®¾ö¾È±¤
È£³áÈò
¾çÇüÁ¤
±è¼öÇü
À̱ͻó
¿À¾Æ¶õ
Anh-Quang Duong
Ngoc-Huynh Ho
Hyung-Jeong Yang
Soo-Hyung Kim
Gueesang Lee
A Ran Oh
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 48 NO. 01 PP. 0709 ~ 0711 (2021. 06) |
Çѱ۳»¿ë (Korean Abstract) |
|
¿µ¹®³»¿ë (English Abstract) |
Emotion recognition in the wild faces with large variances in input signals, multiple sources of noise that will challenge the machine to learn approximate ground truth. In this paper, we propose a jointbased fusion model, called deep graph fusion, to leverage the combination of visual-audio representations for evoked prediction of viewers' expression from videos. We evaluate the proposed method using the Evoked Expression from Videos (EEV) dataset and our experimental results demonstrate that our proposed algorithm outperforms baseline models. |
Å°¿öµå(Keyword) |
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|