COMBINATION OF HANDCRAFTED AND DEEP LEARNING-BASED FEATURES FOR 3D MESH QUALITY ASSESSMENT
| Affiliation auteurs | !!!! Error affiliation !!!! |
| Titre | COMBINATION OF HANDCRAFTED AND DEEP LEARNING-BASED FEATURES FOR 3D MESH QUALITY ASSESSMENT |
| Type de publication | Conference Paper |
| Year of Publication | 2020 |
| Auteurs | Abouelaziz I, Chetouani A, Hassouni MEl, Latecki LJan, Cherifi H |
| Conference Name | 2020 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP) |
| Publisher | Inst Elect & Elect Engineers; Inst Elect & Elect Engineers, Signal Proc Soc |
| Conference Location | 345 E 47TH ST, NEW YORK, NY 10017 USA |
| ISBN Number | 978-1-7281-6395-6 |
| Mots-clés | blind mesh visual quality assessment, curvature, dihedral angle, Fine-tuning, statistical distribution |
| Résumé | We propose in this paper a novel objective method to evaluate the perceived visual quality of 3D meshes. The proposed method in no-reference, it relies only on the distorted mesh for the quality estimation. It is based on a pre-trained convolutional neural network (i.e VGG to extract features from the distorted mesh) and handcrafted features extracted directly from the 3D mesh (i.e curvature and dihedral angle). A General Regression Neural Network (GRNN) is used to learn the statistical parameters of the feature vectors and estimate the quality score. Experimental results from for subjective databases (LIRIS masking, LIRIS/EPFL general-purpose, UWB compression and LEETA simplification) and comparisons with objective metrics cited in the state-of-the-art demonstrate the efficacy of the proposed metric in terms of the correlation to the mean opinion scores across these databases. |