How to run?

[ryanhz]

./darknet detector demo cfg/coco.data cfg/yolov2.cfg yolo.weights -prefix 0

Give me error message saying like cannot load image "data/labels/126_1.png"
then showing a static image. What am I missing?

[SmitSheth]

@huzhiren Can you take pull of the repository. I have updated the repo.

[ryanhz]

Thanks. But after git pull, the screen got stuck at the very beginning of the video.

[SmitSheth]

@huzhiren Are you trying to run the code on GPU/CPU. If on GPU turn on the GPU and CUDNN in Makefile. Set them to 1 instead of 0 and then run make command again

[ryanhz]

Hi @SmitSheth I tried turn on GPU and CUDNN in Makefile, then I got this error:
make: *** [obj/convolutional_layer.o] Error 1 In file included from ./src/utils.h:5:0, from ./src/gemm.c:2: include/darknet.h:11:30: fatal error: cuda_runtime.h: No such file or directory #include "cuda_runtime.h"
Tried to install cuda following this: https://gist.github.com/ingo-m/60a21120f3a9e4d7dd1a36307f3a8cce, but failed.

Turned off GPU, but turned on CUDMN results in:
gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DCUDNN -Wall -Wno-unused-result -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DCUDNN -c ./src/convolutional_layer.c -o obj/convolutional_layer.o
./src/convolutional_layer.c: In function ‘get_workspace_size’:
./src/convolutional_layer.c:91:9: warning: implicit declaration of function ‘cudnnGetConvolutionForwardWorkspaceSize’ [-Wimplicit-function-declaration]
cudnnGetConvolutionForwardWorkspaceSize(cudnn_handle(),
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
./src/convolutional_layer.c:91:49: warning: implicit declaration of function ‘cudnn_handle’ [-Wimplicit-function-declaration]
cudnnGetConvolutionForwardWorkspaceSize(cudnn_handle(),
^~~~~~~~~~~~
./src/convolutional_layer.c:92:18: error: ‘layer {aka struct layer}’ has no member named ‘srcTensorDesc’
l.srcTensorDesc,
^
compilation terminated due to -Wfatal-errors.
Makefile:89: recipe for target 'obj/convolutional_layer.o' failed
make: *** [obj/convolutional_layer.o] Error 1

I am running it from the latest Debian stretch.

[SmitSheth]

@huzhiren Refer the official Nvidia site for installation of CUDA and CUDNN- https://developer.nvidia.com/designworks

You will have to install CUDNN library first before using it. Just for testing purpose you can turn off both GPU and CUDNN flag in Makefile and use yolov2 since its lightweight and will run on CPU as well

[AfreenZehra]

I have installed yolov3 it runs very slow on CPU as I don't have GPU, as well as this is not running with my usb camera