Beatrice Rossi

Spectral Synchronization of Multiple Views in SE(3)

This paper addresses the problem of rigid-motion synchronization (a.k.a. motion averaging) in the Special Euclidean Group SE(3), which finds application in structure-from-motion and registration of multiple three-dimensional (3D) point-sets. After relaxing the geometric constraints of rigid motions, we derive a simple closed-form solution based on a spectral decomposition, which is then projected onto SE(3). Our formulation is extremely efficient, as rigid-motion synchronization is cast to an eigenvalue decomposition problem. Robustness to outliers is gained through Iteratively Reweighted Least Squares. Besides providing a theoretically appealing solution, since our method recovers at the same time both rotations and translations, we demonstrate through experimental results that our approach is significantly faster than the state of the art, while providing accurate estimates of rigid motions.

Global Registration of 3D Point Sets via LRS Decomposition

This paper casts the global registration of multiple 3D point-sets into a low-rank and sparse decomposition problem. This neat mathematical formulation caters for missing data, outliers and noise, and it benefits from a wealth of available decomposition algorithms that can be plugged-in. Experimental results show that this approach compares favourably to the state of the art in terms of precision and speed, and it outperforms all the analysed techniques as for robustness to outliers.

On Computing the Translations Norm in the Epipolar Graph

This paper deals with the problem of recovering the unknown norm of relative translations between cameras based on the knowledge of relative rotations and translation directions. We provide theoretical conditions for the solvability of such a problem, and we propose a two-stage method to solve it. First, a cycle basis for the epipolar graph is computed, then all the scaling factors are recovered simultaneously by solving a homogeneous linear system. We demonstrate the accuracy of our solution by means of synthetic and real experiments.

ECG Monitoring in Wearable Devices by Sparse Models

Because of user movements and activities, heartbeats recorded from wearable devices typically feature a large degree of variability in their morphology. Learning problems, which in ECG monitoring often involve learning a user-specific model to describe the heartbeat morphology, become more challenging.

Robust and Efficient Camera Motion Synchronization via Matrix Decomposition

In this paper we present a structure-from-motion pipeline based on the synchronization of relative motions derived from epipolar geometries. We combine a robust rotation synchronization technique with a fast translation synchronization method from the state of the art. Both reduce to computing matrix decompositions: low-rank & sparse and spectral decomposition. These two steps successfully solve the motion synchronization problem in a way that is both efficient and robust to outliers. The pipeline is global for it considers all the images at the same time. Experimental validation demonstrates that our pipeline compares favourably with some recently proposed methods.

Uncalibrated dynamic stereo using parallax

In this paper a novel method for computing parallax maps from monocular and uncalibrated video sequences is described. Acquired frames are processed pairwise, starting from a first reference and progressively integrating information coming from subsequent frames in temporal order using a Kalman filter. In this way, temporal stabilization of the generated maps is obtained as well as more consistency with the real video content. Results and evidences coming from the benchmark of the system on both synthetic and natural images have also been reported, showing significant improvements with respect to the parallax maps obtained without temporal integration.

Uncalibrated View Synthesis with Homography Interpolation

This paper presents a novel approach to uncalibrated view synthesis that overcomes the sensitivity to the epipole of existing methods. The approach follows a interpolate-then-derectify scheme, as opposed to the previous derectify-then-interpolate strategy. Both approaches generate a trajectory in an uncalibrated framework that is related to a specific Euclidean counterpart, but our method yields a warping map that is more resilient to errors in the estimate of the epipole, as it is confirmed by synthetic experiments.

Packetizing scalable streams in heterogenus peer-to-peer networks

After the extensive effort dedicated by both Academia and Industry in the area of peer-to-peer (P2P) streaming by looking at models and network algorithms to achieve optimal load distribution, recent works are moving forward to enhance the overall P2P systems efficiency by focusing on specific video codecs and packetization techniques at application layer. The purpose to integrate different technologies with the aim to design full streaming solutions requires a joint efforts from the market and EU projects community. Many engineering issues have been brought out in terms of compatibility and integration that need to be addressed. In this paper we focused on a bottleneck discovered during the packetization step between the Scalable Video Coding (SVC) streaming module (during content creation) and the P2P engine before delivering the packets over the network. We compared the a priori fixed packet size solution adopted in P2P-Next project with a codec aware approach, gaining performance in terms of network overhead. We performed experiments with various sequences aiming at overall P2P streaming efficiency and measured the bandwidth cost under various conditions. Since we focused on the application layer, our statistics analysis can be helpful in designing P2P streaming solutions dealing with any network type.

A combined color-correlation visual model for object tracking using particle filters

In this paper we consider the problem of tracking semi-rigid objects in video sequences using particle filters, with a particular focus on hand tracking applications. Although many different feature descriptors have been developed, none of them alone is good enough to deal with this complex tracking scenarios. Approaches which use a statistical representation of the target tend to fail in presence of visually similar objects, while holistic representations are usually very sensible to motion blur, object deformations and rotations. We present here a visual model which combines color histograms and the MOSSE Correlation Filter. The fusion of two complementary features creates a robust descriptor of the target which is capable of tracking fast moving objects in complex tracking applications with real-time performances, using a low number of particles.

Smart 3D video coding

In this paper we explore a method to encode/decode multiview textures and associated supplementary data (depth map as an example) together in a single bitstream. Least significant bit (LSB) of last two levels in each transform block has been modified according to the depth encoded bits, thus resulting in very minimal loss in PSNR of textures (0.1 - 0.5dB) with overall bitrate gain of about 13% for natural scenes. The remaining encoded information have been inserted after the slice NAL of corresponding texture ensuring backward compatibility and correct extraction and decoding by a modified decoder. Benchmarking results in terms of compression efficiency and quality on HD sequences have been reported to explore further the viability of using this technique for Free Viewpoint Video (FVV) or 3DV use cases. Experiments are also carried out to evaluate subjective quality of the synthesized views.