(1) Have the largest control of the generated motion (2) Can be used to generate motion for animals besides human beings
(1) Laborious and time consuming and require the animator to have rich experience
Physics-based methods
(1) Significantly decrease the time of manual adjustment (2) Generate improved results in mechanical and strong regular motion (3) Guarantee the motion in accordance with the physical law (4) Can be used to generate motion for animals besides human beings
(1) Difficult to use when producing smooth and emotional movements, such as dancing and sneaking (2) Generate motion in accordance with the physical law but are not natural and real (3) Entail high computational complexity (4) Feature a physical controller that is difficult to construct
Video-based methods
(1) Require simple data acquisition (2) Entail low cost equipment
(1) Require a single background for acquisition (2) Characterized by poor reusability of synthetic motion data (3) Extract movement data from videos with lower accuracy than motion capture data
Motion capture data-driven methods
(1) Can generate realistic and smooth motion (2) Feature low computational complexity (3) Showcase improved reusability of synthetic motion data
(1) Entail high costs for motion capture equipment (2) Generally applied only to human motion generation
