A Similarity Measure for Material Appearance

Abstract

We present a model to measure the similarity in appearance between different materials, which correlates with human similarity judgments. We first create a database of 9,000 rendered images depicting objects with varying materials, shape and illumination. We then gather data on perceived similarity from crowdsourced experiments; our analysis of over 114,840 answers suggests that indeed a shared perception of appearance similarity exists. We feed this data to a deep learning architecture with a novel loss function, which learns a feature space for materials that correlates with such perceived appearance similarity. Our evaluation shows that our model outperforms existing metrics. Last, we demonstrate several applications enabled by our metric, including appearance-based search for material suggestions, database visualization, clustering and summarization, and gamut mapping.

arXiv comment12 pages, 17 figures

References in corpus (11)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• FaceNet: A Unified Embedding for Face Recognition and Clustering
• UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
• How transferable are features in deep neural networks?
• Proceedings of the 29th International Conference on Machine Learning (ICML-12)
• On the Convergence of Adam and Beyond
• Learning Multi-modal Similarity
• An intuitive control space for material appearance
• Gaussian Material Synthesis
• A Lightweight Approach for On-the-Fly Reflectance Estimation
• Learning icons appearance similarity

Cited by in corpus (8)

• The joint role of geometry and illumination on material recognition
• Material Measurement Units for a Circular Economy: Foundations through a Review
• How Will It Drape Like? Capturing Fabric Mechanics from Depth Images
• Single-image Full-body Human Relighting
• A Unification Between Deep-Learning Vision, Compartmental Dynamical Thermodynamics, and Robotic Manipulation for a Circular Economy
• Tool Substitution with Shape and Material Reasoning Using Dual Neural Networks
• Deep Learning for Material recognition: most recent advances and open challenges
• Not All the Same: Understanding and Informing Similarity Estimation in Tile-Based Video Games