Quality-of-visual-experience: perceptual assessment, compression and enhancement

近年来，数字图像/视频采集、传输和显示技术取得了显著进步。这些技术旨在为消费者提供增强的视觉内容消费体验，并正在为新的多媒体应用程序带来巨大变化。为了在快速增长的全球市场中保持和提高竞争力，加拿大学术界和工业界必须保持和加强在推进研究前沿和开发尖端技术方面的领导地位。该项目旨在探索视觉体验质量(QOVE)研究的理论和应用前沿。首先，我们将开发适用于网络视觉通信的新的客观QoVE评估方法；其次，我们将开发适用于三维(3D)应用环境的客观QoVE评估技术；第三，我们将开发新的压缩和质量增强方法，以优化最佳的2D和3D QoVE。为了实现这一目标，我们将探索一系列具有挑战性的问题-我们将通过整合QoVE模型来开发感知视频压缩方法；我们将在最先进的稀疏表示、压缩传感和统计图像建模方法中融入QoVE标准，以实现视频信号的去噪、超分辨率和帧速率自适应；我们还将研究基于PolyView融合的方法，以及网络视觉通信中的前后处理方法，以进一步改善最终用户的视觉体验。

Recent years have witnessed remarkable advances in digital image/video acquisition, transmission and display technologies. These technologies aim to offer consumers with enhanced experience in consuming visual content, and are creating a vast change in novel multimedia applications. To maintain and increase the competitiveness in the rapidly growing global market, it is critical for Canadian academia and industry to maintain and enhance their leadership in advancing the research fronts and in developing cutting-edge technologies. The proposed project aims to explore the theoretical and application frontiers in quality-of-visual-experience (QoVE) research. First, we will develop novel objective QoVE assessment methodologies suited for network visual communications; Second, we will develop objective QoVE evaluation techniques for three-dimensional (3D) application environment; Third, we will develop novel compression and quality enhancement methodologies optimized for the best 2D and 3D QoVE. To achieve this goal, we will exploit a series of challenging problems - We will develop perceptual video compression methodologies by incorporating QoVE models; We will incorporate QoVE criteria into state-of-the-art sparse representation, compressed sensing, and statistical image modeling methodologies for the denoising, superresolution, and frame rate adaption of video signals; We will also investigate polyview fusion-based methods, and pre- and post-processing methods in network visual communications to further improve the visual experience of end users.