The color information are the most commonly used prior knowledge for depth map super-resolution (DSR), which can provide high-frequency boundary guidance for detail restoration. However, its role and functionality in DSR have not been fully developed. In this article, we rethink the utilization of color information and propose a hierarchical color guidance network (HCGNet) to achieve DSR. On the one hand, the low-level detail embedding (LDE) module is designed to supplement high-frequency color information of depth features in a residual mask manner at the low-level stages. On the other hand, the high-level abstract guidance (HAG) module is proposed to maintain semantic consistency in the reconstruction process by using a semantic mask that encodes the global guidance information. The color information of these 2-D plays a role in the front and back ends of the attention-based feature projection (AFP) module in a more comprehensive form. Simultaneously, the AFP module integrates the multiscale content enhancement (MCE) block and adaptive attention projection (AAP) block to make full use of multiscale information and adaptively project critical restoration information in an attention manner for DSR. Compared with the state-of-the-art methods on four benchmark datasets, our method achieves more competitive performance both qualitatively and quantitatively. The code and results can be found from the link of https://rmcong.github.io/HCGNet_TIM2024 .

中文翻译：

---

学习深度图超分辨率的分层颜色指导

颜色信息是深度图超分辨率（DSR）最常用的先验知识，可以为细节恢复提供高频边界指导。然而，它在 DSR 中的作用和功能尚未得到充分开发。在本文中，我们重新思考颜色信息的利用，并提出了一种分层颜色引导网络（HCGNet）来实现 DSR。一方面，低级细节嵌入（LDE）模块旨在在低级阶段以残差掩模的方式补充深度特征的高频颜色信息。另一方面，提出了高级抽象指导（HAG）模块，通过使用编码全局指导信息的语义掩码来保持重建过程中的语义一致性。这些2D的颜色信息以更全面的形式在基于注意力的特征投影（AFP）模块的前后端发挥作用。同时，AFP模块集成了多尺度内容增强（MCE）模块和自适应注意力投影（AAP）模块，以充分利用多尺度信息，并以注意力方式自适应地为DSR投影关键恢复信息。与四个基准数据集上最先进的方法相比，我们的方法在定性和定量上都实现了更具竞争力的性能。代码和结果可以从以下链接找到https://rmcong.github.io/HCGNet_TIM2024 。