COMMUNITY

BOARD

News

Paper Accepted at 'WACV 2024', a Premier Conference in Computer Vision

Author
College of Software Convergence
Date
2023-12-08
View
27

Files

The research team of Professor Kang Suk-ju, Department of Electronic Engineering,

Paper Accepted at 'WACV 2024', a Premier Conference in Computer Vision


▲ (Top row, from left) Professor Kang Suk-ju, Department of Electronic Engineering; Yang Chang-hee, master's student

(Bottom row, from left) Kang Chan-hee, master's student; Professor Kong Kyeong-bo, Pusan National University (a Sogang graduate)

 

A paper presented by the research team of Professor Kang Suk-ju of the Department of Electronic Engineering (master's student Yang Chang-hee of Electronic Engineering, master's student Kang Chan-hee of the Department of Artificial Intelligence (DHE), and Professor Kong Kyeong-bo of Pusan National University) has been finally accepted at WACV 2024, a premier conference in computer vision.

 

The paper, titled 'Human Motion Aware Text-to-Video Generation with Explicit Camera Control', is the first to propose a model that generates a human skeleton corresponding to the input text and then generates a video containing a person on the basis of that skeleton. It was recognized for its excellence, having been designed to overcome the limitations of models generating videos of people performing complex movements.

 

Kang Chan-hee, a master's student who took part in the research, said: "It is a great honour to have had a WACV publication confirmed during my first semester of graduate school, and I hope it motivates Sogang graduate students and everyone in our laboratory. I hope many more of our students will submit papers to international conferences and go on to see them published."


▲ The model architecture proposed in the paper

 


▲ The three problems in video generation models addressed by the paper


▲ Comparison of results with and without the paper's method applied

 

WACV 2024, organized by IEEE/CVF, will be held in Waikoloa, Hawaii, from 4 to 8 January 2024.

 

 

▶ Paper title: Human Motion Aware Text-to-Video Generation with Explicit Camera Control


Source: Sogang People https://www.sogang.ac.kr/research/res_01.html