JGyoung33 / ARCNN-FAST

ARCNN-FAST tensorflow implementation

ARCNN-FAST tensorflow

[Deep Convolution Networks for Compression Artifacts Reduction] (https://arxiv.org/pdf/1608.02778.pdf)

ARCNN model architecture

ARCNN result (original, jpg20, reconstructed by arcnn on jpg20)

loss history

Prerequisites

• python 3.x
• Tensorflow > 1.5
• matplotlib
• argparse
• opencv2

Properties (what's different from reference code)

• This code requires Tensorflow. This code was fully implemented based on Python 3
• This implements supports color channels based on YCbCr.
• This implements reached PSRN score over paper baseline (on Y channel)

Author

Dohyun Kim