Can Learned Frame-Prediction Compete with Block-Motion Compensation for Video Coding?

TL;DR

This paper explores whether learned frame prediction can rival traditional block-motion compensation in video coding, showing promising results with a simple, pre-trained model that outperforms x264 but not x265 in rate-distortion.

Contribution

It demonstrates that learned frame prediction can achieve competitive compression performance without motion side information, highlighting its potential as an alternative to traditional methods.

Findings

Learned frame prediction outperforms x264 in rate-distortion on test videos.

The simple codec does not yet reach x265 performance.

Training loss functions significantly affect prediction and compression efficiency.

Abstract

Given recent advances in learned video prediction, we investigate whether a simple video codec using a pre-trained deep model for next frame prediction based on previously encoded/decoded frames without sending any motion side information can compete with standard video codecs based on block-motion compensation. Frame differences given learned frame predictions are encoded by a standard still-image (intra) codec. Experimental results show that the rate-distortion performance of the simple codec with symmetric complexity is on average better than that of x264 codec on 10 MPEG test videos, but does not yet reach the level of x265 codec. This result demonstrates the power of learned frame prediction (LFP), since unlike motion compensation, LFP does not use information from the current picture. The implications of training with L1, L2, or combined L2 and adversarial loss on prediction performance and compression efficiency are analyzed.