vot visual object tracking Pose Estimation 3d reconstruction Computer Vision qwen2.5-vl vlm DUSt3R Video Instance Segmentation Video Object Segmentation 3D VISION Calibration videollama3 videollama qwen mcitrack dam4sam sam 2 aaai mast3r structure-from-motion 3D GS solvePnP Projective Geometry Gaussian Splatting Video Segmentation Video Detection Instance Perception CVPR 2023 COLMAP sfm depth estimation 2023 Image Segmentation Image detection CVPR Mamba Didi localization Samurai DRM Dam