Abstract
Feature analysis is a fundamental research area in computer graphics; meanwhile, meaningful and part-aware feature bases are always demanding. This paper proposes a framework for conducting feature analysis on a three-dimensional (3D) model by introducing modified Non-negative Matrix Factorization (NMF) model into the graphical feature space and push forward further applications. By analyzing and utilizing the intrinsic ideas behind NMF, we propose conducting the factorization on feature matrices constructed based on descriptors or graphs, which provides a simple but effective way to raise compressed and scale-aware descriptors. In order to enable part-aware model analysis, we modify the NMF model to be sparse and constrained regarding to both bases and encodings, which gives rise to Sparse and Constrained Non-negative Matrix Factorization (SAC-NMF). Subsequently, by adapting the analytical components (including hidden variables, bases, and encodings) to design descriptors, several applications have been easily but effectively realized. The extensive experimental results demonstrate that the proposed framework has many attractive advantages, such as being efficient, extendable, and so forth.
Highlights
Feature space analysis has, all the time, attracted broad attention, due to its fundamental role in assisting graphical tasks, such as shape understanding, recognition, decomposition, etc. [1,2,3]
Inspired by the Non-negative Matrix Factorization (NMF)-driven applications in computer vision [13,14,15,16] and the related work to factorize the affinity matrix in graphics [17], we propose encoding the 3D model with part-aware bases that afford visually meaningful tools for further applications
We introduce Sparse and Constrained Non-negative Matrix Factorization (SAC-NMF) onto the feature matrix, which may be extracted from one single model or a pair of models, depending on specific applications
Summary
All the time, attracted broad attention, due to its fundamental role in assisting graphical tasks, such as shape understanding, recognition, decomposition, etc. [1,2,3]. We propose factorizing the feature space of three-dimensional (3D) models into meaningful and part-aware hidden variables and bases to facilitate further graphical applications. Face recognition) in computer vision [13,14,15,16] and the related work to factorize the affinity matrix in graphics [17], we propose encoding the 3D model with part-aware bases that afford visually meaningful tools for further applications. We introduce Sparse and Constrained Non-negative Matrix Factorization (SAC-NMF) onto the feature matrix, which may be extracted from one single model or a pair of models, depending on specific applications This kind of feature space analysis can undoubtedly enable joint analysis across the models concerned. 2. We introduce the SAC-NMF to achieve sparse and part-aware analytical components (bases, encodings, and hidden variables) for feature analysis. We adapt analytical components to construct descriptors to empower various applications, including symmetry detection, correspondence, segmentation, and saliency detection
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have