Publications

2022

• GIPSO: Geometrically Informed Propagation for Online Adaptation in 3D LiDAR Segmentation
  
  • Saltori Cristiano
  • Krivosheev Evgeny
  • Lathuilière Stéphane
  • Sebe Nicu
  • Galasso Fabio
  • Fiameni Giuseppe
  • Ricci Elisa
  • Poiesi Fabio
  , 2022. 3D point cloud semantic segmentation is fundamental for autonomous driving. Most approaches in the literature neglect an important aspect, i.e., how to deal with domain shift when handling dynamic scenes. This can significantly hinder the navigation capabilities of self-driving vehicles. This paper advances the state of the art in this research field. Our first contribution consists in analysing a new unexplored scenario in point cloud segmentation, namely Source-Free Online Unsupervised Domain Adaptation (SF-OUDA). We experimentally show that state-of-the-art methods have a rather limited ability to adapt pre-trained deep network models to unseen domains in an online manner. Our second contribution is an approach that relies on adaptive self-training and geometric-feature propagation to adapt a pre-trained source model online without requiring either source data or target labels. Our third contribution is to study SF-OUDA in a challenging setup where source data is synthetic and target data is point clouds captured in the real world. We use the recent SynLiDAR dataset as a synthetic source and introduce two new synthetic (source) datasets, which can stimulate future synthetic-to-real autonomous driving research. Our experiments show the effectiveness of our segmentation approach on thousands of real-world point clouds. Code and synthetic datasets are available at https://github.com/saltoricristiano/gipso-sfouda.
• Outline and Shape Reconstruction in 2D
  
  • Ohrhallinger Stefan
  • Parakkat Amal Dev
  • Peethambaran Jiju
  , 2022. Outline and shape reconstruction from unstructured points in a plane is a fundamental problem with many applications that have generated research interest for decades. Involved aspects like handling open, sharp, multiple and non-manifold outlines, run-time and provability, and potential extension to 3D for surface reconstruction have led to many different algorithms. This multitude of reconstruction methods with quite different strengths and focus makes it difficult for users to choose a suitable algorithm for their specific problem. In this tutorial, we present proximity graphs, graph-based algorithms, and algorithms with sampling guarantees, all in detail. Then, we show algorithms targeted at specific problem classes, such as reconstructing from noise, outliers, or sharp corners. Examples of the evaluation will show how its results can guide users in selecting an appropriate algorithm for their input data. As a special application, we show the reconstruction of lines in the context of sketch completion and sketch simplification. Shape characterization of dot patterns will be shown as an additional field closely related to boundary reconstruction.
• Secure and Robust MIMO Transceiver for Multicast Mission Critical Communications
  
  • Jagyasi Deepa
  • Coupechoux Marceau
  IEEE Transactions on Vehicular Technology, Institute of Electrical and Electronics Engineers, 2022, 71 (6), pp.6351-6366. Mission-critical communications (MCC) involve all communications between people in charge of the safety of the civil society. MCC have unique requirements that include improved reliability, security and group communication support. In this paper, we propose secure and robust Multiple-Input-Multiple-Output (MIMO) transceivers, designed for multiple Base Stations (BS) supporting multicast MCC in presence of multiple eavesdroppers. We formulate minimization problems with the Sum-Mean-Square-Error (SMSE) at legitimate users as an objective function, and a lower bound for the MSE at eavesdroppers as a constraint. Security is achieved thanks to physical layer security mechanisms, namely MIMO beamforming and Artificial Noise (AN). Reliability is achieved by designing a system which is robust to two types of channel state information errors: stochastic and norm-bounded. We propose a coordinate descent-based algorithm and a worst-case iterative algorithm to solve these problems. Numerical results at physical layer and system level reveal the crucial role of robust designs for reliable MCC. We show the interest of both robust design and AN to improve the security gap. We also show that full BS cooperation in preferred for highly secured and reliable MCC but dynamic clustering allows to trade-off security and reliability against capacity. Index Terms-mission critical communication (MCC), physical layer security, robust transceiver design I. INTRODUCTION Mission critical communications (MCC) are all communications between people in charge of the security and the safety of the civil society. Employees of public safety services, like policemen, firemen, rescue teams and ambulance nurses, but also from large companies managing critical infrastructures in the energy or transportation sectors require MCC for their operations [1]. MCC are conveyed by dedicated Private Mobile Radio (PMR) networks [2] that offer a group (or multicast) communication service. This is a one-to-many or many-tomany communication [3], which is one of the most important features of PMR networks and is essential to manage teams of employees. In 5G New Radio, group communication will be supported for MCC from Release R17 onwards [4]. Due to the critical aspects of their missions, MCC users also inherently require highly reliable and secure communication. In particular, sensitive information should not leak to unintended receivers although the broadcast nature of the wireless channel makes the network vulnerable to malicious eavesdroppers. (10.1109/TVT.2022.3160348)
  DOI : 10.1109/TVT.2022.3160348
• On the decoding of lattices constructed via a single parity check
  
  • Corlay Vincent
  • Boutros Joseph J
  • Ciblat Philippe
  IEEE Transactions on Information Theory, Institute of Electrical and Electronics Engineers, 2022, 68 (5), pp.2951 - 2968. This paper investigates the decoding of a remarkable set of lattices: We treat in a unified framework the Leech lattice in dimension 24, the Nebe lattice in dimension 72, and the Barnes-Wall lattices. A new interesting lattice is constructed as a simple application of single parity-check principle on the Leech lattice. The common aspect of these lattices is that they can be obtained via a single parity check or via the king construction. We exploit these constructions to introduce a new efficient paradigm for decoding. This leads to efficient list decoders and quasi-optimal decoders on the Gaussian channel. Both theoretical and practical performance (point error probability and complexity) of the new decoders are provided. (10.1109/TIT.2022.3148196)
  DOI : 10.1109/TIT.2022.3148196
• High resolution neural texture synthesis with long range constraints
  
  • Gonthier Nicolas
  • Gousseau Yann
  • Ladjal Saïd
  International Journal of Mathematical Imaging and Vision(JMIV), 2022, 64, pp.478-492. The field of texture synthesis has witnessed important progresses over the last years, most notably through the use of Convolutional Neural Networks. However, neural synthesis methods still struggle to reproduce large scale structures, especially with high resolution textures. To address this issue, we first introduce a simple multi-resolution framework that efficiently accounts for long-range dependency. Then, we show that additional statistical constraints further improve the reproduction of textures with strong regularity. This can be achieved by constraining both the Gram matrices of a neural network and the power spectrum of the image. Alternatively one may constrain only the autocorrelation of the features of the network and drop the Gram matrices constraints. In an experimental part, the proposed methods are then extensively tested and compared to alternative approaches, both in an unsupervised way and through a user study. Experiments show the interest of the multi-scale scheme for high resolution textures and the interest of combining it with additional constraints for regular textures.
• Escaping limit cycles: Global convergence for constrained nonconvex-nonconcave minimax problems
  
  • Pethick Thomas
  • Latafat Puya
  • Patrinos Panagiotis
  • Fercoq Olivier
  • Cevher Volkan
  , 2022. This paper introduces a new extragradient-type algorithm for a class of nonconvex-nonconcave minimax problems. It is well-known that finding a local solution for general minimax problems is computationally intractable. This observation has recently motivated the study of structures sufficient for convergence of first order methods in the more general setting of variational inequalities when the so-called weak Minty variational inequality (MVI) holds. This problem class captures non-trivial structures as we demonstrate with examples, for which a large family of existing algorithms provably converge to limit cycles. Our results require a less restrictive parameter range in the weak MVI compared to what is previously known, thus extending the applicability of our scheme. The proposed algorithm is applicable to constrained and regularized problems, and involves an adaptive stepsize allowing for potentially larger stepsizes. Our scheme also converges globally even in settings where the underlying operator exhibits limit cycles. Moreover, a variant with stochastic oracles is proposed-making it directly relevant for training of generative adversarial networks. For the stochastic algorithm only one of the stepsizes is required to be diminishing while the other may remain constant, making it interesting even in the monotone setting.