Binarized High-Efficiency RAW Video Restoration and Beyond
2026-08-17 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors present BinRVR, a method to improve video quality from RAW footage using a lightweight binary neural network that is much faster and smaller than usual but only slightly less accurate. They introduce two new techniques: one that helps the model understand both space and time together, and another that reduces errors from simplifying the network's calculations. Their approach also allows adjusting the detail level for different devices. Tests show their method works well for various video improvements like brightening, removing noise, sharpening, and zooming, and it also helps with tasks like detecting objects and estimating depth in videos.
RAW video restorationbinary neural networks (BNNs)temporal coherenceBinarized Information Interaction Module (BIIM)Distribution-Aware Binarized Convolution (DAB-Conv)quantizationlow-light enhancementdenoisingdeblurringsuper-resolution
Authors
Tianyu Zhu, Ying Fu, Hesong Li, Gengchen Zhang, Xin Yuan, Yulun Zhang
Abstract
RAW video restoration is fundamental to high-quality low-level perception and serves as the basis for a wide range of downstream vision applications. While binary neural networks (BNNs) enable efficient lightweight deployment for image enhancement, their deficiencies in modeling temporal coherence and activation value distributions hinder their effectiveness when applied to video scenarios. In this paper, we propose BinRVR, a binarized RAW video restoration framework that reduces computation and parameters by approximately 96% while incurring only about 4% performance degradation. Specifically, we present a Binarized Information Interaction Module (BIIM) to jointly model spatial and temporal information in an efficient and unified manner. Moreover, we develop a Distribution-Aware Binarized Convolution (DAB-Conv) that leverages the statistics of full-precision activations to mitigate quantization errors. The proposed framework further supports multi-bit quantization, enabling flexible accuracy-efficiency trade-offs across different hardware constraints. Extensive experiments demonstrate that our BinRVR achieves competitive performance compared with state-of-the-art binarized methods on RAW video restoration tasks, including low-light enhancement, denoising, deblurring, and super-resolution. We further explore the potential of our method on downstream video applications, including object detection and monocular depth estimation.