About
27
Publications
7,865
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
192
Citations
Publications
Publications (27)
Low-Rank Adaptation (LoRA) methods have gained popularity in efficient parameter fine-tuning of models containing hundreds of billions of parameters. In this work, instead, we demonstrate the application of LoRA methods to train small-vision models in Federated Learning (FL) from scratch. We first propose an aggregation-agnostic method to integrate...
A lot of recent progress has been made in ultra low-bit quantization, promising significant improvements in latency, memory footprint and energy consumption on edge devices. Quantization methods such as Learned Step Size Quantization can achieve model accuracy that is comparable to full-precision floating-point baselines even with sub-byte quantiza...
Machine Learning (ML) has become state of the art for various tasks, including classification of accelerometer data. In the world of Internet of Things (IoT), the available hardware with low-power consumption is often microcontrollers. However, one of the challenges for embedding machine learning on microcontrollers is that the available memory spa...
Deep neural networks are the state of the art in many computer vision tasks. Their deployment in the context of autonomous vehicles is of particular interest, since their limitations in terms of energy consumption prohibit the use of very large networks, that typically reach the best performance. A common method to reduce the complexity of these ar...
Deep neural networks are the state of the art in many computer vision tasks. Their deployment in the context of autonomous vehicles is of particular interest, since their limitations in terms of energy consumption prohibit the use of very large networks, that typically reach the best performance. A common method to reduce the complexity of these ar...
Structured pruning is a popular method to reduce the cost of convolutional neural networks, that are the state of the art in many computer vision tasks. However, depending on the architecture, pruning introduces dimensional discrepancies which prevent the actual reduction of pruned networks. To tackle this problem, we propose a method that is able...
Convolutional neural networks (CNN) have proven very effective in a variety of practical applications involving Artificial Intelligence (AI). However, the layer depth of CNN deepens as user applications become more sophisticated, resulting in a huge number of operations and increased memory size.
The massive amount of the produced intermediate data...
Introduced in the late 1980s for generalization purposes, pruning has now become a staple for compressing deep neural networks. Despite many innovations in recent decades, pruning approaches still face core issues that hinder their performance or scalability. Drawing inspiration from early work in the field, and especially the use of weight decay t...
Polar codes can theoretically achieve very competitive Frame Error Rates. In practice, their performance may depend on the chosen decoding procedure, as well as other parameters of the communication system they are deployed upon. As a consequence, designing efficient polar codes for a specific context can quickly become challenging. In this paper,...
Introduced in the late 80's for generalization purposes, pruning has now become a staple to compress deep neural networks. Despite many innovations brought in the last decades, pruning approaches still face core issues that hinder their performance or scalability. Drawing inspiration from early work in the field, and especially the use of weight-de...
Flexibility is one mandatory aspect of channel coding in modern wireless communication systems. Among other things, the channel decoder has to support several code lengths and code rates. This need for flexibility applies to polar codes that are considered for control channels in the future 5G standard. This paper presents a new generic and flexibl...
AFF3CT is an open source toolbox dedicated to Forward Error Correction (FEC or channel coding). It supports a broad range of codes: from widespread turbo codes and Low-Density Parity-Check (LDPC) codes to more recent polar codes. The toolbox is written in C++ and can be used either as a simulator to quickly evaluate algorithms characteristics, or a...
Dans cet article nous présentons un environne-ment de simulation de Monte Carlo pour les systèmes de communications numériques. Nous nous focalisons en particulier sur les fonctions associées au codage de canal. Après avoir présenté les enjeux liés à la simulation , nous identifions trois problèmes inhérents à ce type de simulation. Puis nous prése...
The recent evolution of mobile communication systems toward a 5G network is associated with the search for new types of non-orthogonal modulations such as Sparse Code Multiple Access (SCMA). Such modulations are proposed in response to demands for increasing the number of connected users. SCMA is a non-orthogonal multiple access technique that offe...
Les codes polaires constituent une classe de codes correcteurs d’erreurs inventés récemment qui suscite l’intérêt des chercheurs et des industriels, comme en atteste leur sélection pour le codage des canaux de contrôle dans la prochaine génération de téléphonie mobile (5G). Un des enjeux des futurs réseaux mobiles est la virtualisation des traiteme...
Cloud Radio Access Network is foreseen as one of the key features of the future 5G mobile communication standard. In this context, all the baseband processing is intended to be performed on CPUs in order to keep a high level of flexibility. The challenge is then to propose efficient software implementation of baseband processing algorithms that gua...
This demonstration intends to present AFF3CT (A Fast Forward 3rror Correction Tool). The main objective of AFF3CT is to provide a portable, open source, fast and flexible software to the channel coding community in such a way that researchers can spend more time on channel coding / algorithmic problems instead of software development issues. It is...