Daniel Kurnia Suhendro

@ilazakis yes I am running this on my laptop.

@abhigoku10, @yossibiton, @azuryl, I have modified DCNv2 from this repository to add the CPU functionality. I have submitted a pull request to Charles Shang, but so far there is no...

Hi, @macqueen09 @abhigoku10, Charles Shang has merged my repo with his, now DCNv2 in this repo can operate using cpu or gpu. Because of this, I do not need to...

@abhigoku10 I have deleted DCNv2 from my repo. Check again this link https://github.com/CharlesShang/DCNv2 readme. it now has a line that says run python testcpu.py to check if it runs on...

@tabsun Hi, I am happy to hear the CPU implementation works for you. Thanks for sharing about openmp too. I was going to suggest that you try making a CPU...

@rshivansh Yes, the code is very slow. If you want to optimize it, you have to create a lookup table for the offsets at each layer of your model.

I do not have the source code for the tensorflow version. The only source code I know is the one provided by the authors of the CFL paper: https://github.com/cfernandezlab/CFL

I used a guide from huggingface (https://medium.com/huggingface/from-tensorflow-to-pytorch-265f40ef2a28) to convert TF version of the trained model and weights to PyTorch. This guide will not provide how to create a working training...

I cannot explain the benefits myself. I can only point you to the part of the paper on robustness analysis. It appears EquiConvs is more robust against distortion of the...