hdrnet-mobile

Introduction

A C++ implementation of inference part of the paper: Gharbi, Michaël, et al. "Deep bilateral learning for real-time image enhancement." ACM Transactions on Graphics (TOG) 36.4 (2017): 118.

Usage

1. create binary model dir from tensorflow ( pretrained models )

python script/get_tfmodel_value.py

2. build shared library and c++ executable

make -B

3. test a picture

python script/run.py A_PICTURE(like xxx.jpg) your_binary_model_dir

To be done

1. optimize the c++ code
2. create an Android NDK version
3. use OpenCL to accelerate
4. retrain the hdrnet+ model, change it to a post process tensorflow model

Reference

1. MIT hdrnet's webpage
2. Gharbi's hdrnet_legacy(forked)
3. Pretrained models