Valeria Garro

Solving the PnP Problem with Anisotropic Orthogonal Procrustes Analysis

In this paper we formulate the Perspective-n-Point (a.k.a. exterior orientation) problem in terms of an instance of the an isotropic orthogonal Procrustes problem, and derive its solution. Experiments with synthetic and real data demonstrate that our method reaches the best trade-off between speed and accuracy. The MATLAB code reported in the paper testifies that it is also exceedingly simple to implement.

Shape Retrieval of Non-rigid 3D Human Models

Abstract3D models of humans are commonly used within computer graphics and vision, and so the ability to distinguish between body shapes is an important shape retrieval problem. We extend our recent paper which provided a benchmark for testing non-rigid 3D shape retrieval algorithms on 3D human models. This benchmark provided a far stricter challenge than previous shape benchmarks. We have added 145 new models for use as a separate training set, in order to standardise the training data used and provide a fairer comparison. We have also included experiments with the FAUST dataset of human scans. All participants of the previous benchmark study have taken part in the new tests reported here, many providing updated results using the new data. In addition, further participants have also taken part, and we provide extra analysis of the retrieval results. A total of 25 different shape retrieval methods are compared.

A Novel Interpolation Scheme for Range Data with Side Information

Time-of-Flight matrix sensors currently available allow for the acquisition of range maps at video rate but usually have a limited resolution. At the same time high resolution color cameras are widely available. This makes highly desirable methods that are able to exploit the combined use of ToF sensors and color cameras to obtain high resolution range maps. This work presents a novel interpolation technique that exploits side information from a standard color camera to increase the resolution of range maps. A segmented version of the high resolution color image is used in order to identify the main objects in the scene, while a novel surface prediction scheme is used to interpolate the available depth samples. Critical issues like the joint calibration of the two devices and the unreliability of the acquired data have also been taken into account with ad- hoc solutions. The performance of the proposed scheme has been verified with both synthetic and real-world data and experimental results have shown how the proposed method allows to obtain a more accurate interpolation with sharper edges if compared with standard approaches.

Shape retrieval of non-rigid 3D human models

We have created a new benchmarking dataset for testing non-rigid 3D shape retrieval algorithms, one that is much more challenging than existing datasets. Our dataset features exclusively human models, in a variety of body shapes and poses. 3D models of humans are commonly used within computer graphics and vision, and so the ability to distinguish between body shapes is an important shape retrieval problem. In this track nine groups have submitted the results of a total of 22 different methods which have been tested on our new dataset.

Omnidirectional image capture on mobile devices for fast automatic generation of 2.5D indoor maps

We introduce a light-weight automatic method to quickly capture and recover 2.5D multi-room indoor environments scaled to real-world metric dimensions. To minimize the user effort required, we capture and analyze a single omni-directional image per room using widely available mobile devices. Through a simple tracking of the user movements between rooms, we iterate the process to map and reconstruct entire floor plans. In order to infer 3D clues with a minimal processing and without relying on the presence of texture or detail, we define a specialized spatial transform based on catadioptric theory to highlight the rooms structure in a virtual projection. From this information, we define a parametric model of each room to formalize our problem as a global optimization solved by Levenberg-Marquardt iterations. The effectiveness of the method is demonstrated on several challenging real-world multi-room indoor scenes.

Edge-preserving interpolation of depth data exploiting color information

The extraction of depth information associated to dynamic scenes is an intriguing topic, because of its perspective role in many applications, including free viewpoint and 3D video systems. Time-of-flight (ToF) range cameras allow for the acquisition of depth maps at video rate, but they are characterized by a limited resolution, specially if compared with standard color cameras. This paper presents a super-resolution method for depth maps that exploits the side information from a standard color camera: the proposed method uses a segmented version of the high-resolution color image acquired by the color camera in order to identify the main objects in the scene and a novel surface prediction scheme in order to interpolate the depth samples provided by the ToF camera. Effective solutions are provided for critical issues such as the joint calibration between the two devices and the unreliability of the acquired data. Experimental results on both synthetic and real-world scenes have shown how the proposed method allows to obtain a more accurate interpolation with respect to standard interpolation approaches and state-of-the-art joint depth and color interpolation schemes.

Retrieval and classification methods for textured 3D models: a comparative study

This paper presents a comparative study of six methods for the retrieval and classification of textured 3D models, which have been selected as representative of the state of the art. To better analyse and control how methods deal with specific classes of geometric and texture deformations, we built a collection of 572 synthetic textured mesh models, in which each class includes multiple texture and geometric modifications of a small set of null models. Results show a challenging, yet lively, scenario and also reveal interesting insights into how to deal with texture information according to different approaches, possibly working in the CIELab as well as in modifications of the RGB colour space.

Fast metric acquisition with mobile devices

We present a novel algorithm for fast metric reconstruction on mobile devices using a combination of image and inertial acceleration data. In contrast to previous approaches to this problem, our algorithm does not require a long acquisition time or intensive data processing and can be implemented entirely on common IMU-enabled tablet and smartphones. The method recovers real world units by comparing the acceleration values from the inertial sensors with the ones inferred from images. In order to cope with IMU signal noise, we propose a novel RANSAC-like strategy which helps to remove the outliers. We demonstrate the effectiveness and the accuracy of our method through an integrated mobile system returning point clouds in metric scale.

Multiscale descriptors and metric learning for human body shape retrieval

The aim of this paper was to show the usefulness of applying feature projection or metric learning techniques to multiscale descriptor spaces for the effective retrieval of human bodies of labeled subjects. Using learned subspace projections it is possible to strongly improve the retrieval performance obtained with state-of-the-art global descriptors, and, in some cases, to perform an effective feature fusion. Results obtained on different human scan datasets show that Linear Discriminant Analysis, applied to Histograms of Area Projection Transform and Shape DNA features after a preliminary dimensionality reduction, creates compact descriptors that are quite effective in improving the subject retrieval scores both when class (subject) examples are available in the training set and when only examples of classes not included in the test set are used for training. Other mappings tested are less effective even if still able to improve the results. Retrieval scores obtained in the same experimental settings used in recent related papers show that the approach based on our mapped features largely outperforms the other methods proposed for the task, even those specifically designed for human body characterization.

TreeSha: 3D shape retrieval with a tree graph representation based on the autodiffusion function topology

In this paper we present a new method for shape description and matching based on a tree representation built upon the scale space analysis of maxima of the Autodiffusion function (ADF). The use of the Heat Kernel based approach makes the method invariant to articulated deformations. By coupling maxima of the Autodiffusion function with the related basins of attraction, it is possible to link the information at different scales encoding spatial relationships in a tree structure. Furthermore, texture information can be easily included in the descriptor by adding regional color histograms to the node attributes of the tree. Dedicated graph kernels have been designed to evaluate shape dissimilarity from the obtained representations using both structural, geometric and color information. Preliminary experiments performed on the SHREC 2013 non-rigid textured dataset showed very good retrieval performances.