Current Result Document :
| 한글제목(Korean Title) |
트랜스포머의 효과적인 시간 특징 정보 학습을 위한 합성곱 기법 |
| 영문제목(English Title) |
Convolutional Approach to Learning Temporal Feature Effectively in Transformer |
| 저자(Author) |
박해성
정혁철
최용석
Hae Sung Park
Hyuck Chul Jung
Yong Suk Choi
|
| 원문수록처(Citation) |
VOL 49 NO. 02 PP. 0517 ~ 0519 (2022. 12) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
In the video classification task, a well-performing deep learning model is likely to extract proper temporal features to classify the data. However, we found out several problems of attention-based TimeSformer[2] related to extracting temporal features, and replaced the time attention module in the TimeSformer with the 3D convolution module for better temporal feature processing. Through several experiments and visualization results, we demonstrate that the 3D convolution module can extract more accurate temporal features of video data than the time-attention module. |
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
