Fig. 7From: Collision-aware interactive simulation using graph neural networksQualitative evaluation. The proposed method is evaluated in three different collision scenes: vertex-face, edge-face, and face-face collisions. For each scene, the first row is the simulation result of the proposed method, and the second row is the ground truth. a: Vertex-face collisions; b: Edge-face collisions; c: Face-face collisionsBack to article page