ifzhang
Processing frame 0 (100000.00 fps)
RuntimeError: No CUDA GPUs are available [root@localhost ByteTrack-main]# python3.6 tools/demo_track.py video -f exps/example/mot/yolox_x_mix_det.py -c pretrained/bytetrack_x_mot17.pth.tar --save_result Error in cpuinfo: processor architecture is not supported in cpuinfo 2022-02-22 09:10:46.165 | INFO | main:main:316 - Args: Namespace(aspect_ratio_thresh=1.6, camid=0, ckpt='pretrained/bytetrack_x_mot17.pth.tar', conf=None, demo='video', device=device(type='cuda'), exp_file='exps/example/mot/yolox_x_mix_det.py', experiment_name='yolox_x_mix_det', fp16=False, fps=30, fuse=False, match_thresh=0.8, min_box_area=10, mot20=False, name=None, nms=None, path='./videos/palace.mp4', save_result=True, track_buffer=30, track_thresh=0.5, trt=False, tsize=None) 2022-02-22 09:11:22.329 | INFO | main:main:326 - Model Summary: Params: 99.00M, Gflops: 791.73 2022-02-22 09:11:22.376 | INFO | main:main:334 - loading checkpoint 2022-02-22 09:11:28.400 | INFO | main:main:338 - loaded checkpoint done. 2022-02-22 09:11:28.417 | INFO | main:imageflow_demo:248 - video save_path is ./YOLOX_outputs/yolox_x_mix_det/track_vis/2022_02_22_09_11_28/palace.mp4 2022-02-22 09:11:28.418 | INFO | main:imageflow_demo:258 - Processing frame 0 (100000.00 fps) 2022-02-22 09:11:28.420 | INFO | main:imageflow_demo:298 - save results to ./YOLOX_outputs/yolox_x_mix_det/track_vis/2022_02_22_09_11_28.txt
unknow no inference ?
That's because of this line.
In the first iteration, since the timer hasn't yet been started (it is activated when running predictor.inference(frame, timer)), the expression 1. / max(1e-5, timer.average_time) will return 100000, because timer.average_time is still 0.
~~No idea~~, I tried changing it, but couldn't get it to work the intended way. EDIT: I took a second look. If you want to just fix it not showing 100000 fps at the start, the easiest fix i would recommend is just changing 0 to 19 here: https://github.com/ifzhang/ByteTrack/blob/2c082be967a994632be254eb3ddbad8280e9e0a6/tools/demo_track.py#L257 Good luck in your endeavours :smiley: