SIGGRAPH 2019์์ ๋ฐํ๋ ๋ ผ๋ฌธ์ธ “Learning Character-Agnostic Motion for Motion Retargeting in 2D”์ ํ๋ก์ ํธ ํ์ด์ง ๊ณต์ ํฉ๋๋ค.
์ด ๋ ผ๋ฌธ์ (์๋ก ๋ฌ๋ผ๋ ๋๋) ์์ 3๊ฐ๋ก๋ถํฐ ๊ฐ๊ฐ motion, skeleton, camera angle์ ์ถ์ถํ ํ์ ์ด๋ค์ ํ๋๋ก ๊ฒฐํฉํ๋ ๋ฅ๋ฌ๋ ๋ชจ๋ธ์ ๋ง๋ค์์ต๋๋ค. Motion, Skeleton, Camera Angle ๊ฐ๊ฐ์ ๋ํ ๊ฐ๋ณ, ํน์ ์กฐํฉ ๋ฆฌํ๊ฒํ ์ด ๊ฐ๋ฅํฉ๋๋ค. ํ์ต๋ฐ์ดํฐ๋ก๋ Mixamo๋ฅผ ์ฌ์ฉํ์ต๋๋ค.
๊ฒฐ๊ณผ๋ฌผ์ video performance cloning์ ๋ฌผ๋ก motion retrieval ๋ฑ ๋ค์ํ ๋ถ์ผ์์ ํ์ฉ ๊ฐ๋ฅํ๋ค๊ณ ์ ์๊ฐ ๋ฐํ๊ณ ์์ต๋๋ค. ์ ๋ ํฅํ 3D ํน์ 2D ์บ๋ฆญํฐ ์ ๋๋ฉ์ด์ ์ ํ์ฉํด ๋ณผ ์๊ฐ์ ๋๋ค.
Motion Retargeting in 2D
Analyzing human motion is a challenging task with a wide variety of applications in computer vision and in graphics.
One such application, of particular importance in computer animation, is the retargeting of motion from one performer to another.
While humans move in three dimensions, the vast majorโฆ
One such application, of particular importance in computer animation, is the retargeting of motion from one performer to another.
While humans move in three dimensions, the vast majorโฆ