MMdnn is a set of tools to help users inter-operate among different deep learning frameworks. E.g. model conversion and visualization. Convert models between Caffe, Keras, MXNet, Tensorflow, CNTK, PyTorch Onnx and CoreML.
MIT License
5.79k
stars
964
forks
source link
Adding support for Spatial Pyramid Pooling layer #295
I have a model trained in caffe which uses SPP layer. I was trying to convert this but there is no support for an SPP layer in MMdnn. Since this layer uses pooling and the shape and stride of the operation is determined on runtime, I was wondering how I could add support for this. Furthermore, there is no direct support for an SPP layer in tensorflow either.
I have a model trained in caffe which uses SPP layer. I was trying to convert this but there is no support for an SPP layer in MMdnn. Since this layer uses pooling and the shape and stride of the operation is determined on runtime, I was wondering how I could add support for this. Furthermore, there is no direct support for an SPP layer in tensorflow either.
Any pointers would be really helpful.