Kuan Wang

Thanks :) This repo is just the draft for this project, and the newest version is in my another repo opencv_contrib(branch gsoc_opencv). The final version will be part of OpenCV...

remember check out to branch gsoc_opencv :)

you can refer to the BilateralSolver.cpp of the repo The_Bilateral_Solver to build the depth demo with bilateralsolver. And the Lattice filter is not a part of this OpenCV moudle, you...

Thanks for your interest. Yes, the linear quantization part will be released soon.

We have updated the linear quantization part in this repo. Please let us know if you have any questions.

We have updated the linear quantization as well as the hardware resource-constrained part in this repo. Please let us know if you have any questions.

Thanks for your feedback! You can view the related code refer to https://github.com/mit-han-lab/haq/blob/7141586e9ae47c8a50aa8b596ab37682a06b434a/lib/env/linear_quantize_env.py#L306

[QConv2d, QLinear] are linear quantization layers. And you can find the implementation in https://github.com/mit-han-lab/haq/blob/7141586e9ae47c8a50aa8b596ab37682a06b434a/lib/utils/quantize_utils.py#L121

The paper updated once on 14 May, but still has some bugs. I think there should not be non-linear after pool.

Thanks a lot! I've just updated the codes. The gap of the model size is still large, welcome to report more bugs :)