Yolo v5 not showing Object detection Prediction Results with GPU on RISC-V.

[siri243]

Hello Team,

We've tried to build the NCNN Framework on our RISC-V Development Platform, Where we're seeing the Yolo v5 Example Test Application is providing approximate test prediction values on CPU but in addition to this we're also noticing "find_blob_index_by_name failed" logged in the test output. When we tried to verify the same Test Application on the same platform with Imagination GPU we're not seeing any prediction values to be outputted along with "find_blob_index_by_name failed" error.

1) Without GPU: vkEnumeratePhysicalDevices failed -3 find_blob_index_by_name 353 failed Try ex.extract("output", out0); ex.extract("376", out1); ex.extract("401", out2); find_blob_index_by_name 367 failed Try ex.extract("output", out0); ex.extract("376", out1); ex.extract("401", out2); 49 = 0.73544 at 351.99 489.19 134.91 x 114.68 49 = 0.69195 at 630.90 476.09 140.60 x 127.97 terminate called after throwing an instance of 'cv::Exception' what(): OpenCV(4.5.4) ./modules/highgui/src/window_gtk.cpp:635: error: (-2:Unspecified error) Can't initialize GTK backend in function 'cvInitSystem'

2) With GPU module: find_blob_index_by_name 353 failed Try ex.extract("output", out0); ex.extract("376", out1); ex.extract("401", out2); find_blob_index_by_name 367 failed Try ex.extract("output", out0); ex.extract("376", out1); ex.extract("401", out2); terminate called after throwing an instance of 'cv::Exception'

Can you please help us resolving the same. Any help is highly appreciated Thank you!

Regards, Siri Ravinath

[wzyforgit]

It seems that you have a wrong input and output blob name. Check your param file and your source code.

[siri243]

Sure we'll check on this, thank you