search

ChenYingpeng / caffe-yolov3

A real-time object detection framework of Yolov3/v4 based on caffe
474 stars 231 forks source link

can train & test own dataset after yolov3 transfer to caffe model ? #44

Closed Ashing00 closed 4 years ago

Ashing00 commented 4 years ago

basically, I use "./x86_64/bin/detectnet 0 ../../data/yolov3/prototxt/yolov3.prototxt ../../data/yolov3/caffemodel/yolov3.caffemodel" to inference original yolov3 model for coco dataset is ok the dog.jpg have the correct box. but when I train my dataset and do inference ,the result is strange .a lot of boxes appear .my dataset only 10 class . below picture is the test result.

Ashing00 commented 4 years ago

I test it on original Yolov3 by darknet is ok,but when transfer the model to caffe and inference it is strange.a lot box appear.please help me ,thank you.

hand

Ashing00 commented 4 years ago

sorry,I already found out the solution ,thanks all.

RatkoFri commented 3 years ago

@Ashing00 Could you share some detail on how did you adjust YOLO network for your problem?

Ratko