tiny yolo v3 inerence is slow one core cpu no gpu

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )

http://pjreddie.com/darknet/

Other

21.68k stars 7.96k forks source link

tiny yolo v3 inerence is slow one core cpu no gpu #5765

Open Usernamezhx opened 4 years ago

Usernamezhx commented 4 years ago

first of all. thanks for your cool work. I run the darknet_video.py . and it give me inference time about 14s I change darknet_video.py to print time not fps. one frame. Is this normal?

my cpu Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz my Cmake:

this is my darknet_video.py file

Usernamezhx commented 4 years ago

OPENMP=0 AVX=1 1.155052661895752 s OPENMP=0 AVX=0 5.507767677307129 s OPENMP=1 AVX=0 17.84518837928772 s

Usernamezhx commented 4 years ago

I will test this https://github.com/AlexeyAB/darknet/wiki/Converting-Yolo-v3-models-to-TensorFlow-and-OpenVINO(IR)-models

AlexeyAB commented 4 years ago

Just use yolov3-tiny with OpenCV-dnn without Darknet: https://docs.opencv.org/master/da/d9d/tutorial_dnn_yolo.html

Also compile OpenCV (with IE OpenVINO backend): https://github.com/opencv/opencv/wiki/Intel's-Deep-Learning-Inference-Engine-backend