Inria Rennes - Bretagne Atlantique Research Centre

This study aims to identify the stroke mechanics regulations that underlie pacing optimization and differentiate final performance in the 400 m front-crawl across different performance levels. 157 all-out 400 m trials from trained to elite swimmers equipped with a sacrum worn IMU were analyzed. Hierarchical generalized additive models were fitted to describe (1) the technical predictors of final performance, through the interrelationships between averaged kinematics (i.e., stroke rate (SR), stroke length (SL), jerk cost (JC)) and mean speed, and (2), the profiles of pacing and associated stroke mechanics regulations leading to performance. These models depict both the common patterns shared by all swimmers and the specific group deviations by performance level (G1 -slowest- to G4 -fastest-) from population trends. All swimmers shared a reverse-J-shaped pacing (p < 0.001). The fastest swimmers reduce pacing variability with a relatively slower start, a minimized drop in speed in the middle of the race, and a greater end-spurt than the population trend (p < 0.05). Their stroke mechanics are effectively adapted to the requirements of such specific race sections, both through a U-shaped JC with higher magnitude (p < 0.001) and a continuous adaptation of the SR (p < 0.01) and SL (p < 0.01) combination. A high average SL (i.e., at least 1.50 m.stroke-1), coupled with an SR close to 70 strokes.min-1 and maximizing stroke smoothness at race pace, are critical technical abilities for optimizing final performance. Technical training should prioritize the reduction of intra-cyclic acceleration variations at race pace as a key driver of 400 m performance development. The best performers lower pacing variability by amplifying the magnitude of kinematic regulations, particularly in terms of stroke smoothness management. Such underlying adaptations in stroke mechanics play a pivotal role in continuous speed control throughout the race. The findings of this study enhance the technical understanding of discriminative 400 m pacing skills for coaches and swimmers.

In a 2023 Modelica Conference paper, we proposed a novelmethod for the modular structural analysis of DAE systems,in which the structural analysis is not performed onflattened models, but rather at the class level. A newnotion of structural interface was proposed, in whichclasses are enriched with context information. That paperdeveloped our approach based on a few illustrative examples.In this paper, we provide the details of our algorithm. Itsperformance depends on the system architecture: theanalysis of models having a small number of classes (possi-bly instantiated many times), with a low treewidth systemarchitecture, scales up very efficiently with thisapproach. We then present additional benchmarks, amongwhich a urban heating network, a representative real-lifeexample on which a near-logarithmic scaling up is shown.

Inverse problems involve reconstructing clean images from degraded observations. Maximum a Posteriori (MAP) estimation reconstructs the most probable source image from noisy measurements. When combined with Plug-and-Play (PnP) priors defined by an image denoising algorithm, MAP estimation yields high-quality reconstructions. In contrast, Diffusion Models (DMs) address inverse problems by sampling from the posterior distribution using score functions trained on images perturbed by Gaussian noise. Prior work reformulated diffusion sampling as Deep Equilibrium (DEQ) models but did not fine-tune DMs for inverse problems. This work introduces MaximUm a PostEriori Training (MUPET), a framework that leverages PnP gradient descent to enable DEQ fine-tuning of DMs on inverse problems. By refining a generative prior at the fixed-point of MAP estimation, MUPET enhances image restoration via posterior sampling while maintaining quality when sampling from the prior.

Reproducibility is fundamental to scientific research, but can be particularly challenging in research that involves High Performance Computing (HPC) due to the unique characteristics of supercomputers.Performance-based metrics such as execution time, energy consumption, and throughput further complicate reproducibility, especially on shared systems.In this paper, we present our experience implementing a reproducibility initiative at SC24, with particular emphasis on changes made compared to prior SC conferences.We outline HPC-specific challenges, describe the measures adopted to address them, and reflect on the limitations of reproducibility badges.Faced with the constraints of the existing badging nomenclature, we discuss our implementation of a reproducibility report, which aims to provide more context about the reproducibility of each paper.We conclude by recommending that the "Artifact Replicable" badge be dropped by HPC conferences at this time, and discuss alternate ways of ensuring replicability evaluation. CCS Concepts General and reference Validation.

Conventional federated learning (FL) coordinated by a central server focuses on training a global model and protecting the privacy of clients' training data by storing it locally. However, the statistical heterogeneity hinders the global model from adapting to the non-IID distributions among clients. Moreover, untrusted and unreliable central servers and malicious clients may compromise model integrity and availability, thus degrading the robustness of FL. To address these challenges, we present RobustPFL, a decentralized personalized federated learning (PFL) approach that combines <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\alpha$</tex-math></inline-formula>-based Layer-position Normalized Similarity (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\alpha$</tex-math></inline-formula>-LNS) and local collaborative training to improve personalized performance while utilizing a blockchain-based committee mechanism to coordinate the aggregation process, thereby achieving high personalized accuracy and robustness. Extensive experiments show that our RobustPFL approach outperforms multiple algorithms, including Local training, FedAvg, FedReptile, Per-FedAvg, FedBN, and SPFL, on MNIST, CIFAR10, EMNIST, and N-BaIoT datasets in four non-IID settings. We also evaluate RobustPFL's effectiveness against attacks—poisoning attacks and free-riding attacks. Particularly, for three prevalent poisoning attacks (backdoor, label flipping, and model poisoning attacks), we compare non-defensive (FedAvg) and defensive (Krum, trimmed mean, Bulyan, FedBN, FLAME, and FangTrmean) methods with our proposed RobustPFL. The results show that our approach achieves significant defensive effects.

Video-based high-density crowd analysis and prediction has been a long-standing topic in computer vision. It is notoriously difficult due to, but not limited to, the lack of high-quality data and complex crowd dynamics. Consequently, it has been relatively under-studied. In this paper, we propose a new approach that aims to learn from in-the-wild videos, often with low quality where it is difficult to track individuals or count heads. The key novelty is a new physics prior to model crowd dynamics. We model high-density crowds as active matter, a continuum with active particles subject to stochastic forces, named ‘crowd material’. Our physics model is combined with neural networks, resulting in a neural stochastic differential equation system that can mimic complex crowd dynamics. Due to the lack of similar research, we adapt a range of existing methods which are close to ours for comparison. Through exhaustive evaluations, we show our model outperforms existing methods in analyzing and forecasting extremely high-density crowds. Furthermore, since our model is a continuous-time physics model, it can be used for simulation and analysis, providing strong interpretability. This is categorically different from most deep learning methods, which are discrete-time models and black-boxes.

Virtual machines (VMs) raise several challenges for efficiently accommodating Cloud-gaming needs. The first challenge concerns VM boot time. Booting a VM can take several milliseconds to several seconds based on the operating system (OS) and the modules that must be loaded. Consequently, waiting for a VM to boot upon a user’s game-start request impacts the game startup latency and, thus, user experience. To mitigate this startup issue, Cloud-gaming providers maintain a pool of pre-initialized VMs, which leads to resource waste when the pool is not fully used. The second challenge concerns the flexibility of VMs regarding their memory size due to the lack of memory hotplug and unplug features. VMs cannot adapt their memory size to accommodate the heterogeneous memory requirements of existing video games. Concretely, pre-initialized VMs for a specific game cannot always be reused to host another game with different memory requirements. This means pre-initialized VMs must be configured with large memory sizes to host several games, leading to memory waste.In this paper, we bootstrap a discussion to steer away from a VM-centric Cloud-gaming infrastructure in favor of more resource-efficient infrastructures. To this end, we take a two-fold approach. First, we propose using containers as an efficient alternative to Windows VMs for video games. Our experiments show that the lightweight nature of containers achieves up to 10.6% energy-usage reduction without impacting the gaming experience. Second, we propose a novel GPU-sharing idea tailored for game containers. We discuss state-of-the-art GPU-sharing approaches, show their practical limitations for Cloud-gaming platforms, and present our envisioned GPU-sharing approach.

International audience