happynear / happynear/caffe-windows
about the code in predict_box_layer.cpp&.cu
Open
- Dominant language
- C++
- Stars
- 1.3k
- Forks
- 637
- PR merge metrics
- No merged PRs in 30d
Description
I read the code and find that when the batch_size or so called num in this layer is bigger than 1, the top[1] is set to all zero. Does it mean this caffe version can not support batch image inference?
Contributor guide
Assessment
This issue has not been assessed yet.