HEVC all Intra Compression Noise Reduction. (Simple example)
use pytorch(python)
- Train step
- Data generation
- set QP and Encoding and frame select. (encoding_data.py)
- Split the data into patches. (createCompressNoisePatch.py)
- convert the data to h5 file (createCompressNoisePatch.py)
- Train Model
- setting training configuration
- train model (train_model.py)
- Data generation
use libtorch(C++)
- serialize pytorch weight file (trace_model.py)
- run trained model in c++ (main.cpp)