search

open-mmlab / mmtracking

OpenMMLab Video Perception Toolbox. It supports Video Object Detection (VID), Multiple Object Tracking (MOT), Single Object Tracking (SOT), Video Instance Segmentation (VIS) with a unified framework.
https://mmtracking.readthedocs.io/en/latest/
Apache License 2.0
3.55k stars 595 forks source link

KeyError: 'detections' #91

Closed yangxiaoyany closed 3 years ago

yangxiaoyany commented 3 years ago

hello,I met an error when i test the demo.mp4 .Thank you for helping me to solve it.

(mmdet) yxy@yxy:/media/yxy/4TB/ayxy/mmtracking$ python demo/demo_mot.py configs/mot/deepsort/sort_faster-rcnn_fpn_4e_mot17-public-half.py --input demo/demo.mp4 --output mot.mp4 2021-01-13 21:36:14,270 - mmtrack - INFO - load detector from: https://download.openmmlab.com/mmtracking/mot/faster_rcnn/faster-rcnn_r50_fpn_4e_mot17-half-64ee2ed4.pth [ ] 0/8, elapsed: 0s, ETA:Traceback (most recent call last): File "demo/demo_mot.py", line 88, in main() File "demo/demo_mot.py", line 64, in main result = inference_mot(model, img, frame_id=i) File "/media/yxy/4TB/ayxy/mmtracking/mmtrack/apis/inference.py", line 78, in inference_mot data = test_pipeline(data) File "/media/yxy/4TB/ayxy/mmdetection/mmdet/datasets/pipelines/compose.py", line 40, in call data = t(data) File "/media/yxy/4TB/ayxy/mmtracking/mmtrack/datasets/pipelines/loading.py", line 100, in call detections = results['detections']

KeyError: 'detections'

GT9505 commented 3 years ago

The config you used (sort_faster-rcnn_fpn_4e_mot17-public-half.py) needs external detections like public detections in MOT17 challenge. Please use config with private detections (e.g sort_faster-rcnn_fpn_4e_mot17-private-half.py) to run demo_mot.py

yangxiaoyany commented 3 years ago

Thank you.it works