Non-parametric Filtering for Geometric Detail Extraction and Material Representation

Abstract

a) image (b) albedo (c) coarse-scale shading (d) shading detail Figure 1. We decompose an image (a) into three components: (b) albedo, (c) coarse-scale shading, and (d) shading detail. The albedo and coarse-scale shading represent surface color and directional lighting effect. The shading detail image captures fine-scale surface geometry, or material property. AbstractGeometric detail is a universal phenomenon in real world objects. It is an important component in object modeling, but not accounted for in current intrinsic image works. In this work, we explore using a non-parametric method to separate geometric detail from intrinsic image components. We further decompose an image as albedo * (coarse-scale shading + shading detail). Our decomposition offers quantitative improvement in albedo recovery and material classification.Our method also enables interesting image editing activities, including bump removal, geometric detail smoothing/enhancement and material transfer.

Dimensionality Reduction on SPD Manifolds: The Emergence of Geometry-Aware Methods

“…Table 1 compares the performance of the studied algorithms. The performance of the stateof-the-art method on this dataset was reported to be 43.5% Liao et al (2013). The results show that appropriate manifold-based methods (i.e., kSC-S and CDL) with the original 155 × 155 RCMs already outperform this state-of-the-art, while nearest neighbour (e.g., NN-AIRM, NN-S) on the same manifold yields worse performance.…”

“…For the task of material categorization, we used the UIUC dataset Liao et al (2013). The UIUC material dataset contains 18 subcategories of materials taken in the wild from four general categories (see Fig.…”

“…Recently, SIFT features Lowe (2004) have been shown to be robust and discriminative for material classification Liao et al (2013). Therefore, we constructed RCMs of size 155 × 155 using 128 dimensional SIFT features (from grayscale images) and 27 dimensional color descriptors.…”

“…The dataset provides the 3D locations of 31 joints over time acquired at the speed of 120 frames per second. We describe an action of a K joints skeleton observed over m frames by its joint covariance descriptor Hussein et al (2013), Figure 5: Samples from the UIUC material dataset Liao et al (2013). Table 3: Recognition accuracies for the YTC dataset Kim et al (2008).…”

Vectorial approximations of infinite-dimensional covariance descriptors for image classification

“…For the first task, we used the UIUC Material dataset [4], which contains 18 categories of materials taken in the wild. Each category has 12 images of various sizes with various geometric details included.…”