search

tukilabs / Video-Compression-Net

A new approach to video compression by refining the shortcomings of conventional approach and substituting each traditional component with their neural network counterpart. Our proposed work consists of motion estimation, compression and compensation and residue compression, learned end-to-end to minimize the rate-distortion trade off. The whole model is jointly optimized using a single loss function. Our work is based on a standard method to exploit the spatio-temporal redundancy in video frames to reduce the bit rate along with the minimization of distortions in decoded frames. We implement a neural network version of conventional video compression approach and encode the redundant frames with lower number of bit.
25 stars 12 forks source link

I want H.264 vompression evaluation benchmark values #4

Closed durgabhavaniv closed 2 years ago

durgabhavaniv commented 3 years ago

Is that possible to get compression images with traditional h.264 compression and I want to evaluate those results like finding PSNR and MS-SSIM as shown in https://arxiv.org/pdf/1812.00101.pdf this paper benchmark results.

prasangadhungel commented 3 years ago

Hi @durgabhavaniv ,

We have mentioned the link to our experimental results in the experiment section. In case you haven't checked, here it is. It contains the original UVG dataset, and the frames compressed using h.264, h.265 and ours. PS: We cannot share the JCT-VC dataset for the copyright issues.

Best wishes, Prasanga Dhungel

durgabhavaniv commented 3 years ago

hi @PrasangaDhungel

Thanks for update, Actually I want to compress images with h.264 by my self using traditional compression technique, similar to compress.py proposed here. Is that possible ?

Thank you.

prasangadhungel commented 3 years ago

Hi @durgabhavaniv

Yes, the compression is possible but first you have to convert the video into raw frames. FFmpeg can be of great help on this, or you can simply use our conversion script. After that, you can compressed the extracted raw pngs using the compress.py. You can check our own evaluation here

Best, Prasanga