IVCVSep 27, 2022

Neural Frank-Wolfe Policy Optimization for Region-of-Interest Intra-Frame Coding with HEVC/H.265

arXiv:2209.13210v15 citationsh-index: 21
Originality Incremental advance
AI Analysis

This work addresses region-of-interest coding for video compression, which is incremental as it builds on prior RL methods by improving convergence guarantees.

The paper tackled the problem of bit allocation for region-of-interest intra-frame coding in HEVC/H.265 by introducing a reinforcement learning framework with Frank-Wolfe policy optimization, achieving superior performance compared to baselines in experiments with x265.

This paper presents a reinforcement learning (RL) framework that utilizes Frank-Wolfe policy optimization to solve Coding-Tree-Unit (CTU) bit allocation for Region-of-Interest (ROI) intra-frame coding. Most previous RL-based methods employ the single-critic design, where the rewards for distortion minimization and rate regularization are weighted by an empirically chosen hyper-parameter. Recently, the dual-critic design is proposed to update the actor by alternating the rate and distortion critics. However, its convergence is not guaranteed. To address these issues, we introduce Neural Frank-Wolfe Policy Optimization (NFWPO) in formulating the CTU-level bit allocation as an action-constrained RL problem. In this new framework, we exploit a rate critic to predict a feasible set of actions. With this feasible set, a distortion critic is invoked to update the actor to maximize the ROI-weighted image quality subject to a rate constraint. Experimental results produced with x265 confirm the superiority of the proposed method to the other baselines.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes