search

tusharsangam / TransVisDrone

MIT License
29 stars 3 forks source link

NPS Video/Images Detection Problem #13

Open Stacnia opened 1 month ago

Stacnia commented 1 month ago

Hi Dear Developer,

I have encountered a problem. When I want to perform detection using the already trained model, I found that there is no detect.py file to detect images or videos. Could I be missing any files to perform this task?

Thank you!

tusharsangam commented 2 weeks ago

You can use this python val.py --data ./data/NPS.yaml \ --weights ./runs/train/NPS/image_size_1280_temporal_YOLO5l_5_frames_NPS_end_to_end_skip_0/weights/best.pt \ --batch-size 1 --img 1280 --num-frames 5 \ --project ./runs/val/NPS/image_size_1280_temporal_YOLO5l_5_frames_NPS_end_to_end_skip_0 --name best_augment_full_save \ --task test --exist-ok --save-json --augment --save-json-gt Also in the submit-test.slurm file.