Search

Monocular depth estimation from a single image is an ill-posed problem for computer vision due to insufficient reliable cues as the prior knowledge. Besides the inter-frame supervision, namely stereo and adjacent frames, extensive prior information is available in the same frame. Reflections from specular surfaces, informative intra-frame priors, enable us to reformulate the ill-posed depth estimation task as a multi-view synthesis. This paper proposes the first self-supervision for deep-learning depth estimation on water scenes via intra-frame priors, known as reflection supervision and geometrical constraints. In the first stage, a water segmentation network is performed to separate the reflection components from the entire image. Next, we construct a self-supervised framework to predict the target appearance from reflections, perceived as other perspectives. The photometric re-projection error, incorporating SmoothL1 and a novel photometric adaptive SSIM, is formulated to optimize pose and depth estimation by aligning the transformed virtual depths and source ones. As a supplement, the water surface is determined from real and virtual camera positions, which complement the depth of the water area. Furthermore, to alleviate these laborious ground truth annotations, we introduce a large-scale water reflection scene (WRS) dataset rendered from Unreal Engine 4. Extensive experiments on the WRS dataset prove the feasibility of the proposed method compared to state-of-the-art depth estimation techniques. • Proposes an intra-frame-supervised depth estimation by specular reflections, comprising water segmentation and depth estimation. • Introduces the Photometric Adaptive SSIM for aligning reflections with source patterns, emphasizing local contrast and structural details. • Develops the Water Reflection Scene dataset to address the lack of reflection scenes depth estimation. [ABSTRACT FROM AUTHOR]

Abstract: This paper is a review of the theoretical framework for the application of radiative transfer equations to structural dynamics and acoustics. It is shown that under the assumption of geometrical acoustics and when the phase of rays is neglected, a representation of a sound field in terms of incoherent fictitious sources provides a sufficiently large framework to embody diffuse as well as specular reflection, steady-state or transient phenomena and even diffraction. Two special cases are mentioned. The so-called radiosity method in acoustics corresponds to a purely diffusing boundary and statistical energy analysis when the sound field is diffuse. [Copyright &y& Elsevier]

The multi-layer optical windows are usually installed in spacecraft for scientific observation, in which the homogeneity of temperature distribution has effects on image quality. In this paper, considering semitransparent and specularly reflecting interfaces, based on the ray tracing-node analysis method, a multi-layer transient coupled radiative and conductive heat transfer model is used to examine the effects of thickness and the number of heat insulation glass layers on temperature distributions of the multi-layer optical windows. The changes of absorption coefficient and refractive index of the optical window changing with wave length are approximately expressed by using a set of rectangular spectral bands. The study shows that in outer space, thermal radiation plays a very important role in the cooling process of the multi-layer optical window, and in the very thin layer of glass medium very close to the surface of glass layer, the radiative heat flux and the conductive heat flux are coupled with each other intensively. The thinner the insulation glass layer is, the more homogeneous the temperature distribution therein is, thus helpfully improving the image quality. The more the number of insulation glass layers are, the more homogeneous temperature distribution of the multi-layer optical window is, thus also helpfully improving the image quality. Also determination of the optimized number of heat insulation glass layers needs to consider the other factors. [ABSTRACT FROM AUTHOR]