Hello,Is anyone aware of OpenCL based CNN implementation on Altera FPGAs? I have looked into Intel Deep Learning Inference accelerator (https://software.intel.com/en-us/inference-engine-devguide). It converts models generated by tensorflow and caffe into Intermediate representation. Anyone aware of converting custom graphs in caffe and tensorflow into FPGA binary format (.aocx)? Regards, Manoj.
Check Google Scholars, there are tens of papers on this subject. However, if you are looking for something open-source that is ready to use, you wouldn't find more than one or two.
We recently developed a CNN accelerator based on the OpenCL framework. We used DE10 Nano based on Intel Cyclone V SoC FPGA for the implementation.
You can check out the entire design flow to implement the accelerator and the relevant codes in the following repository: https://github.com/tirumalnaidu/opencl-cnn-accelerator
For more technical information, refer to this presentation: Link