Adaptive High-Frequency Preprocessing for Video Coding
This work addresses bandwidth and storage costs in video coding, but it is incremental as it builds on existing preprocessing and rate-distortion optimization methods.
The paper tackles the trade-off between video clarity and coding bitrate by introducing an adaptive high-frequency preprocessing framework, which enhances subjective quality and saves bitrate, as demonstrated through evaluations on multiple datasets.
High-frequency components are crucial for maintaining video clarity and realism, but they also significantly impact coding bitrate, resulting in increased bandwidth and storage costs. This paper presents an end-to-end learning-based framework for adaptive high-frequency preprocessing to enhance subjective quality and save bitrate in video coding. The framework employs the Frequency-attentive Feature pyramid Prediction Network (FFPN) to predict the optimal high-frequency preprocessing strategy, guiding subsequent filtering operators to achieve the optimal tradeoff between bitrate and quality after compression. For training FFPN, we pseudo-label each training video with the optimal strategy, determined by comparing the rate-distortion (RD) performance across different preprocessing types and strengths. Distortion is measured using the latest quality assessment metric. Comprehensive evaluations on multiple datasets demonstrate the visually appealing enhancement capabilities and bitrate savings achieved by our framework.