Institut Systèmes Intelligents et de Robotique

Diffusion-based image-to-video (I2V) models are increasingly effective, yet they struggle to scale to ultra-high-resolution inputs (e.g., 4K). Generating videos at the model's native resolution often loses fine-grained structure, whereas high-resolution tiled denoising preserves local detail but breaks global layout consistency. This failure mode is particularly severe in the fresco animation setting: monumental artworks containing many distinct characters, objects, and semantically different sub-scenes that must remain spatially coherent over time. We introduce FrescoDiffusion, a training-free method for coherent large-format I2V generation from a single complex image. The key idea is to augment tiled denoising with a precomputed latent prior: we first generate a low-resolution video at the underlying model resolution and upsample its latent trajectory to obtain a global reference that captures long-range temporal and spatial structure. For 4K generation, we compute per-tile noise predictions and fuse them with this reference at every diffusion timestep by minimizing a single weighted least-squares objective in model-output space. The objective combines a standard tile-merging criterion with our regularization term, yielding a closed-form fusion update that strengthens global coherence while retaining fine detail. We additionally provide a spatial regularization variable that enables region-level control over where motion is allowed. Experiments on the VBench-I2V dataset and our proposed fresco I2V dataset show improved global consistency and fidelity over tiled baselines, while being computationally efficient. Our regularization enables explicit controllability of the trade-off between creativity and consistency.

The evolution of colorectal cancer (CRC) follows a progression from polyp to adenoma and eventually to cancer. The implementation of CRC screening is an effective strategy to interrupt the disease process and reduce morbidity and mortality during this process. Early diagnosis and treatment of tumors can be facilitated through widespread screening campaigns, thereby reducing the economic burden on the state, society and families. In order to achieve this goal, the key lies in the development of screening methods that are both efficient and economical to ensure that they can be widely disseminated and minimize the rate of missed diagnoses. At home and abroad, numerous scholars and experts have devoted themselves to the study of CRC screening methods. This review synthesizes current evidence through a systematic literature review to evaluate existing CRC screening methodologies, thereby offering critical insights to inform clinical decision-making in optimizing screening strategy selection.

Unmanned surface vehicles (USVs) have great application prospects in defense, environmental surveillance and offshore energy due to their cost-effectiveness and long-duration mission ability. The structural safety issues induced by the prolonged cyclic wave loading on such small-sized marine structures, such as fatigue failure mechanism, represent an important research topic. In order to characterize the loading process, a piston-type numerical wave flume with wave absorption setting is constructed using the Arbitrary Lagrangian Eulerian (ALE) formulation, and the fluid–structure interaction (FSI) simulations are performed. Simulated wave profiles are measured and compared with corresponding analytical wave solutions to verify the accuracy of target waves. The wave absorption effect is verified by comparing the velocities of water particles in different water regions. Then, different impact scenarios are performed by applying a range of the applicable target waves. Simulated wave forms, impact scenes along with the computed wave load data are presented, and the impact process is analyzed. As a result, the FSI simulations demonstrate cyclic loading characteristics of small-sized floating structures subjected to wave impacts, and the constructed ALE numerical wave flume possesses the extensibility for the simulation of nonlinear water wave impact scenarios.

We investigated whether surface texture (i.e., stochastic roughness) influences softness perception during direct touch interactions with elastic, textured stimuli. Using a Bayesian adaptive modeling approach and a 2AFC task, we evaluated participants' ability to discriminate the softness of stimuli that varied in both their stochastic surface roughness (Hurst exponent) and material elasticity. To explore potential interactions between these features, we conducted two discrimination experiments, testing stimuli from two distinct ranges of elasticity. All participants performed the task using pressing. Results show that softness discrimination was determined primarily by material elasticity, with no discernible influence of surface features. The findings suggest that humans effectively isolate elasticity-based information from smaller-scale surface topography or texture during direct pressing with the finger.

Optical tweezers use tightly focused laser beams to trap and manipulate microscopic particles by balancing scattering forces with a gradient force at the beam’s focal point. This enables precise and contact-free particle control, making optical tweezers a powerful tool to activate microrobots (optobots) that give multiple degrees of freedom. However, realizing out-of-plane rotation, which is demanded for medical applications such as cellmanipulation and drilling, remains a challenge. Here, we present an optobot design leveraging chirality to achieve full-cycle out-of-plane rotation. This optobot has an elongated structure with dual optical handles and a chiral helix aligned on the long axis of the robot. The optical handles are used to be trapped by the optical tweezers beams to keep the long axis of the optobot and the helix perpendicular to the laser light propagation, while its chiral helical part generates out-of-plane rotation around its long axis under laser excitation due to broken axial parity. Finite element analysis was conducted to simulate the interaction of the chiral helix with the laser beam. The scattered beam distribution was visualized to see how it is altered by the chiral structure. The optical torque was also calculated to show that it is unidirectional following chirality. For experimental demonstration, the optobot was fabricated via two-photon lithography. The optical manipulation experiments using an optical tweezer system showed that the optobot demonstrates versatile actuation capabilities, achieving on-demand repetitive rotation, controlled-speed translation, and combined rotational-translational motion.

Many physical tasks require the cooperation of multiple individuals. It is known that during the execution of a joint manipulation task, the forces they exchange carry information. However, how this information is transmitted and interpreted remains partially unclear, necessitating specific devices to increase our knowledge. Even the most technologically advanced devices are generally the result of delicate compromises between the complexity of allowed movements and the controllability of the system’s joints. In this paper, we present the design of the SEMAPHORO-2D, a versatile haptic device with two actuated degrees of freedom. This mechatronic system enables accurate measurement of the forces generated by the users and the creation of a highly rigid and stable bilateral teleoperation link between two partners.

ABSTRACTNonlinearity in parallel compliance can be exploited to improve the performance of locomotion systems in terms of (1) energy efficiency, (2) control robustness, and (3) gait optimality; that is, attaining energy efficiency across a set of motions. Thus far, the literature has investigated and validated only the first two benefits. In this study, we present a new mathematical framework for designing nonlinear compliances in cyclic tasks encompassing all three benefits. We present an optimization‐based formulation for each benefit to obtain the desired compliance profile. Furthermore, we analytically prove that, compared to linear compliance, using nonlinear compliance leads to (1) lower energy consumption, (2) better closed‐loop performance, specifically in terms of tracking error, and (3) a higher diversity of natural frequencies. To compare the performance of linear and nonlinear compliance, we apply the proposed methods to a diverse set of robotic systems performing cyclic tasks, including a 2‐DOF manipulator, a 3‐DOF bipedal walker, and a hopper model. Compared to linear compliance, the nonlinear compliance leads to better performance in all aspects; for example, a 70% reduction in energy consumption and tracking error for the manipulator simulation. Regarding gait optimality, for all robotic simulation models, compared to linear compliance, the nonlinear compliance has lower energy consumption and tracking error over the considered set of motions. The proposed analytical studies and simulation results strongly support the idea that using nonlinear compliance significantly improves robotic system performance in terms of energy efficiency, control robustness, and gait optimality.

Recent advances in AI have led to significant results in robotic learning, but skills like grasping remain partially solved. Many recent works exploit synthetic grasping datasets to learn to grasp unknown objects. However, those datasets were generated using simple grasp sampling methods using priors. Recently, Quality-Diversity (QD) algorithms have been proven to make grasp sampling significantly more efficient. In this work, we extend QDG-6DoF, a QD framework for generating object-centric grasps, to scale up the production of synthetic grasping datasets. We propose a data augmentation method that combines the transformation of object meshes with transfer learning from previous grasping repertoires. The conducted experiments show that this approach reduces the number of required evaluations per discovered robust grasp by up to 20 %. We used this approach to generate QDGset, a dataset of 6 DoF grasp poses that contains about 3.5 and 4.5 times more grasps and objects, respectively, than the previous state-of-the-art. Our method allows anyone to easily generate data, eventually contributing to a large-scale collaborative dataset of synthetic grasps.

Nous examinons dans cet article la contribution des aspects relationnels dans l’efficacité de la prescription de psychotropes chez l’enfant et l’adolescent. Cette composante psychologique est aussi désignée comme « effet placebo » ou « effets non-spécifiques » en recherche clinique. Cette revue narrative synthétise des données épidémiologiques, des résultats d’essais cliniques et des travaux de psychologie expérimentale concernant les facteurs relationnels influençant le niveau de réponse aux psychotropes en population pédiatrique. Les résultats de la littérature ont été intégrés dans un modèle théorique adapté aux spécificités de la prescription en pédopsychiatrie, inspiré par la théorie des systèmes et la psychologie du développement. Ce modèle interactionniste médecin-malade-parent insiste sur les liens entre les représentations préexistantes autour de la prescription (conscientes ou non) entre parents et enfant, et leurs influences sur les croyances et attitudes concrètes vis-à-vis du médicament. Les représentations autour du médicament sont influencées par des facteurs liés aux enjeux développementaux et des expériences de vie parentales, qui peuvent être abordés en consultation pour en moduler les effets. Le prescripteur joue un rôle essentiel pour accueillir ces représentations, aider à élaborer leurs significations. Des exemples cliniques sont présentés pour illustrer la façon dont ces représentations s’expriment au cours des prises en charge. Les recommandations de bonne pratique doivent tenir compte de l’importance des aspects relationnels à toutes les étapes prescription de psychotropes chez l’enfant et l’adolescent pour en maximiser les effets potentiels. This article examines how relational aspects contribute to the effectiveness of prescribing psychotropic drugs to children and adolescents. This psychological component is often referred to as the “placebo effect” or “non-specific effects” in clinical research. This narrative review synthesizes data from epidemiological studies, clinical trials, and experimental psychology regarding the relational factors that influence response levels to psychotropic medications in the pediatric population. The literature findings have been integrated into a theoretical model inspired by systems theory and developmental psychology. This interactionist model emphasizes the connections between the pre-existing perceptions of medication held by parents and children, and how these perceptions influence their beliefs and attitudes toward medication. Representations of medication are shaped by developmental issues and parental life experiences. These factors can be addressed during consultations to mitigate their effects. The prescriber has a crucial role in acknowledging these representations and assisting in their interpretation. Clinical examples are provided to illustrate how these representations manifest during treatment. Guidelines for best practices must consider the significance of relational factors at every stage of prescribing psychotropic medications to children and adolescents to optimize their therapeutic effects.