A Tri-Dynamic Preprocessing Framework for UGC Video Compression

2025-12-17

AI Summary

To address the poor generalization of machine learning-based video coding on user-generated content (UGC) videos—characterized by high spatiotemporal variability—this paper proposes a novel triple-dynamic preprocessing framework. It introduces, for the first time, a synergistic and tunable adaptive mechanism comprising dynamic preprocessing intensity, dynamic quantization level, and dynamic λ-based rate-distortion trade-off. Leveraging a differentiable encoder simulator, adaptive factor modulation, and a dynamically adjusted λ loss function, the framework enables end-to-end joint rate-distortion optimization. Evaluated on a large-scale real-world UGC video benchmark, the method achieves an average BD-rate reduction of 12.6% over state-of-the-art approaches, demonstrating superior compression efficiency while preserving visual quality—thereby striking a more favorable balance between rate and distortion.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In recent years, user generated content (UGC) has become the dominant force in internet traffic. However, UGC videos exhibit a higher degree of variability and diverse characteristics compared to traditional encoding test videos. This variance challenges the effectiveness of data-driven machine learning algorithms for optimizing encoding in the broader context of UGC scenarios. To address this issue, we propose a Tri-Dynamic Preprocessing framework for UGC. Firstly, we employ an adaptive factor to regulate preprocessing intensity. Secondly, an adaptive quantization level is employed to fine-tune the codec simulator. Thirdly, we utilize an adaptive lambda tradeoff to adjust the rate-distortion loss function. Experimental results on large-scale test sets demonstrate that our method attains exceptional performance.

Problem

• Optimizes UGC video compression for diverse content variability
• Enhances data-driven encoding algorithms in UGC scenarios
• Addresses challenges in rate-distortion optimization for UGC videos

Innovation

• Adaptive factor regulates preprocessing intensity
• Adaptive quantization fine-tunes codec simulator
• Adaptive lambda adjusts rate-distortion loss function