A novel method for the classification and retrieval of 3D models is proposed; it exploits the 2D panoramic view representation of 3D models as input to an ensemble of Convolutional Neural Networks which automatically compute the features. The first step of the proposed pipeline, pose normalization is performed using the SYMPAN method, which is also computed on the panoramic view representation. In the training phase, three panoramic views corresponding to the major axes, are used for the training of an ensemble of Convolutional Neural Networks. The panoramic views consist of 3-channel images, containing the Spatial Distribution Map, the Normals' Deviation Map and the magnitude of the Normals' Devation Map Gradient Image. The proposed method aims at capturing feature continuity of 3D models, while simultaneously minimizing data preprocessing via the construction of an augmented image representation. It is extensively tested in terms of classification and retrieval accuracy on two standard large scale datasets: ModelNet and ShapeNet. 1. Introduction 1 In the recent past, convolutional neural networks (CNN) have 2 shown their superiority against humans in computing features, 3 while they are very sensitive to the input representation. In this 4 work an extension of the PANORAMA 3D shape representa-5 tion, previously proposed by our team (Papadakis et al., 2010), 6 is exploited as the input representation to a CNN for computing 7 descriptor features for 3D object classification and retrieval. 8 The 3D models are initially pose normalized using the SYM-9 PAN pose normalization algorithm, (Sfikas et al., 2014) which 10 is based on the use of reflective symmetry on their panoramic 11 view images. Next, an augmented panoramic view is created 12 and used to train the convolutional neural network. This aug-13 mented panoramic view consists of the spatial and orientation 14 components of PANORAMA, (see 3.1.1), along with the mag-15 nitude of the gradient image which is extracted from the ori-16 entation component. A reduction in the size of the augmented 17 panoramic view representation is shown to benefit the training 18 procedure.

A novel pose normalization method based on 3D object reflective symmetry is presented. It is a general purpose global pose normalization method; in this paper it is used to enhance the performance of a 3D object retrieval pipeline. Initially, the axis-aligned minimum bounding box of a rigid 3D object is modified by requiring that the 3D object is also in minimum angular difference with respect to the normals to the faces of its bounding box. To estimate the modified axis-aligned bounding box, a set of predefined planes of symmetry are used and a combined spatial and angular distance, between the 3D object and its symmetric object, is calculated. By minimizing the combined distance, the 3D object fits inside its modified axis-aligned bounding box and alignment with the coordinate system is achieved. The proposed method is incorporated in a hybrid scheme, that serves as the alignment method in a 3D object retrieval system. The effectiveness of the 3D object retrieval system, using the hybrid pose normalization scheme, is evaluated in terms of retrieval accuracy and demonstrated using both quantitative and qualitative measures via an extensive consistent evaluation on standard benchmarks. The results clearly show performance boost against current approaches.

No abstract

This work offers an overview of the state-of-the-art on the emerging area of 3D object retrieval based on partial queries. This research area is associated with several application domains, including face recognition and digital libraries of cultural heritage objects. The existing partial 3D object retrieval methods can be mainly classified as: i) view-based, ii) partbased, iii) bag of visual words (BoVW)-based, and iv) hybrid methods combining these three main paradigms or methods which cannot be straightforwardly classified. Several methodological aspects are identified, including the use of interest points and the exploitation of 2.5D projections, whereas the available evaluation datasets and campaigns are addressed. A thorough discussion follows, identifying advantages and limitations.

No abstract