Open t-m-k opened 6 years ago
your bn layer is different
I looked yolo3-tiny network, it just two outputs.You should changed this code in detectnet.cpp.
I also used yolov3-tiny, but the detection result is not good and different from the result of the original darknet-yolov3-tiny.
First of all, I would like to thank you for your work with passing the Yolov3 to Caffe, it works perfectly. But, as @DeepPower said, the TinyYolo gives me wrong results too (supressing the third blob in the detectnet.cpp).
Hi. I modified the detectnet.cpp and then tested the yolov3-tiny network. But I got an unsatisfactory result. Do you have any good suggestions or ways to change this situation? Thank you.
there is a question when use yolov3-tiny network ,the question is num_inputs is 1 num_outputs is 2 F0807 10:35:53.908385 11051 detectnet.cpp:71] Check failed: net->num_outputs() == 3 (2 vs. 3) Network should have exactly three outputs.
when i modify the outputs in detetcnet.cpp(cancel the thrid output reveice variable,so there are only two feature to calculation result) , the application can show img but there is no effect at all ,even the person in img is so obviously.