🤖 AI Summary
This study addresses the degradation of temporal correlation and compression quality in screen content videos caused by abrupt motion transitions and high-frequency details. To tackle these challenges, this work proposes STM-Net, an enhancement framework that incorporates a prior-guided spatiotemporal scheduler and a parallel stream routing mechanism. These components adaptively handle abrupt transitions without explicit detection while preventing feature contamination. Furthermore, by integrating bidirectional temporal feature extraction with a cascaded multi-scale distillation module, the proposed method effectively preserves critical high-frequency information. Experimental results demonstrate that STM-Net outperforms existing state-of-the-art approaches in both objective metrics and subjective visual quality, offering a robust solution for mitigating compression artifacts in screen content videos.
📝 Abstract
Different from natural videos, Screen Content Videos (SCVs) are characterized by abrupt motion, scene switches, and high-frequency details such as text and graphics. Conventional video enhancement methods, which rely heavily on temporal continuity, often suffer from performance degradation when processing SCVs due to the disruption of temporal correlations. To address these challenges, we propose the Spatial-Temporal Multi-scale Network (STM-Net), a novel framework specifically tailored for compressed SCV enhancement. Our approach integrates three complementary components: a Prior-Guided Spatio-Temporal Dispatcher (PG-STD) that routes input into three parallel streams to avoid feature contamination, a Bidirectional Temporal Feature Extraction (BTFE) module that adaptively handles abrupt transitions without explicit detection, and a Cascaded Multi-scale Feature Distillation (CMFD) module that preserves critical high-frequency details. Experimental results demonstrate that STM-Net outperforms state-of-the-art methods in both objective metrics and subjective visual quality, providing a robust solution for screen content artifacts. Code is available at https://github.com/HUANGZiyin1/STM-Net.