def main(): for object in objects(): #car, truck, pedestrian, ... for idx, file_name in enumerate(file_names): #데이터 한 세트 sequence_mot() -> ids, bboxes, states, types 저장 def sequence_mot(): for frame(): tracker.frame_mot() #tracking ids, bboxes, states, types.append() return ids, bboxes, states, types #모든 frame에 대한 정보가 들어있는 list 반환 class MOTmodel: def frame_mot(): for matched_track: #matched trac..
1. token information raw_data에서 v1.0-trainval 폴더 안에 sample.json에서 token정보를 받아와서 저장. sample.json 예시 { "token": "fd8420396768425eabec9bdddf7e64b6", "timestamp": 1533201470448696, "prev": "", "next": "6eb8a3ff0abf4f3a9380a48f2a0b87ef", "scene_token": "e7ef871f77f44331aefdebc24ec034b7" } token_info/scene-0003.json 예시(token_info.py결과) ["fd8420396768425eabec9bdddf7e64b6", "6eb8a3ff0abf4f3a9380a48f2a0b..
https://arxiv.org/abs/1711.06396 VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection Accurate detection of objects in 3D point clouds is a central problem in many applications, such as autonomous navigation, housekeeping robots, and augmented/virtual reality. To interface a highly sparse LiDAR point cloud with a region proposal network (RP arxiv.org 1. Point cloud의 '어떤 특성'때문에..
submission { "meta": { "use_camera": -- Whether this submission uses camera data as an input. "use_lidar": -- Whether this submission uses lidar data as an input. "use_radar": -- Whether this submission uses radar data as an input. "use_map": -- Whether this submission uses map data as an input. "use_external": -- Whether this submission uses external data as an input. }, "results": { sample_tok..
