PaddlePaddle / PaddleDetection

Object Detection toolkit based on PaddlePaddle. It supports object detection, instance segmentation, multiple object tracking and real-time multi-person keypoint detection.
Apache License 2.0
12.6k stars 2.86k forks source link

how to export txt file?(PP-YOLOE) #6683

Open nomurakeiya opened 2 years ago

nomurakeiya commented 2 years ago

问题确认 Search before asking

请提出你的问题 Please ask your question

thanks for your work!!

I could inference my own images on PP-YOLOE. But, I counldn't export txt file of deetection result. now, I'm using "--save_mot_txt_per_img" or "--save_mot_txts" and arguments is True.

command python deploy/python/infer.py --model_dir=output_inference/ppyoloe_crn_x_300e_coco --image_file=demo/000000014439.jpg --device=GPU --save_mot_txt_per_img

--save_mot_txt_per_img image

--save_mot_txts image

Even this situation, the output is only image. I need txt file to count object. image

this issue is happen PP-YOLO too. please help me.

nemonameless commented 2 years ago

--save_mot_txt_per_img or --save_mot_txts only used in MOT models like FairMOT or ByteTrack. Just a post-processing. You can write simple python codes here

nomurakeiya commented 2 years ago

@nemonameless thanks for your information.

I'm not good at coding, do you have any adea of "simple python codes"?