For a few decades, automatic license plate recognition (ALPR) has been considered as a powerful intelligent video analytics tool to deal with the growing number of vehicles worldwide. Most recent ALPR landmark achievements are contingent on access to diverse and rich datasets for model training and evaluation. In this context, we present a naturalistic dataset of European license plates, named ELP 1.0. Unique to ELP 1.0 is its vast geographical coverage, comprising a wide spectrum of European countries (20) and vehicle categories (10). Using the proposed dataset, we benchmark the performances of several baseline deep learning models, assessing their ability to detect and recognize license plates. Further, to picture the ELP 1.0 dataset’s features, we propose a try-one-subset-out setup and evaluate the generalization abilities of ALPR approaches across an array of scenarios. The experimental findings highlight ELP 1.0 challenges as well as its usefulness to question ALPR models’ capacities. The ELP 1.0 dataset, with its comprehensive annotations, features, and benchmark results, will serve as a testbed to support future ALPR research and, more broadly, intelligent transportation industrial solutions. The ELP 1.0 dataset is accessible upon request via this URL.

Ensuring the safety of workers in hazardous environments is crucial, given the potential severity of accidents. Indeed, Personal Protective Equipment (PPE), including safety helmets, vests, and boots, is vital in mitigating risks. However, adherence to PPE usage often falls short due to inadequate safety awareness among workers. To address this challenge, we aim to develop a robust safety monitoring system by leveraging advancements in deep learning and object detection technology. Our proposed method uses YOLOv7 and YOLOv8 algorithms for real-time detection and classification of safety equipment, including helmets, vests, and boots. For this aim, we carefully collect and annotate datasets from diverse sources like Roboflow and Kaggle. Then, we accurately evaluate performance metrics such as precision, recall, and speed through rigorous training and testing of various YOLOv7 and YOLOv8 configurations on annotated data batches. By conducting thorough comparative analyses, we aim to identify the most effective model for each PPE category, considering both detection accuracy and computational efficiency. Finally, the object detection system is deployed through a web interface to determine whether a person is protected or not.

Automatic license plate recognition (ALPR) is a powerful tool for analyzing the growing number of vehicles in major cities worldwide. However, building datasets that accurately represent vehicles and license plates (LPs) is challenging, as most current datasets are specific to certain countries and do not provide a comprehensive representation. This remains particularly true for French LPs, where there is currently no publicly available dataset that covers a wide range of scenarios such as plate inclinations, varying lighting conditions, and changing weather. To address this issue, we present a benchmark suite called PGFLP 1.0, which includes more than 14k fully annotated LPs from several vehicle categories (e.g. cars, trucks, motorcycles). PGFLP 1.0 includes three annotation levels: rectangular boxes for the LP, four vertices of the LP region, and labels for the full LP sequence. Using PGFLP 1.0, we provide an experimental evaluation of the most recent ALPR approaches, assessing their ability to detect and recognize LPs. This work provides a useful benchmark in terms of accuracy and speed for current and future research efforts in ALPR.

Infrared imaging has recently played an important role in a wide range of applications including video surveillance, robotics and night vision. However, infrared cameras often suffer from some limitations, essentially about low-contrast and blurred details. These problems contribute to the loss of observation of target objects in infrared images, which could limit the feasibility of different infrared imaging applications. In this paper, we mainly focus on the problem of pedestrian detection on thermal images. Particularly, we emphasis the need for enhancing the visual quality of images before performing the detection step. To address that, we propose a novel thermal enhancement architecture called TE-GAN based on Generative Adversarial Network, and composed of two modules contrast enhancement and denoising with a post-processing step for edge restoration in order to improve the overall image quality. The effectiveness of the proposed architecture is assessed by means of visual quality metrics and better results are obtained compared to the original thermal images and to the obtained results by other existing enhancement methods. These results have been conducted on a subset of KAIST dataset that we make available to encourage research in this direction11https://github.com/AmineMarnissi/TE-GAN, Using the same dataset, the impact of the proposed enhancement architecture has been demonstrated on the detection results by obtaining better performance with a significant margin using YOLOv3 detector.

In this paper, we propose a novel idea to address the problem of fast computation of stable force-closure grasp configurations for a multifingered hand and a 3-D rigid object represented as a polygonal soup model. The proposed method performs a low-level shape exploration by wrapping multiple cords around the object in order to quickly isolate promising grasping regions. Around these regions, we compute grasp configurations by applying a variant of the close-until-contact procedure to find the contact points. The finger kinematics and the contact information are then used to filter out unstable grasps. Through many simulated examples with three different anthropomorphic hands, we demonstrate that, compared with previous grasp planners such as the generic grasp planner in Simox, the proposed grasp planner can synthesize grasps that are more natural-looking for humans (as measured by the grasp quality measure skewness) for objects with complex geometries in a short amount of time. Unlike many other planners, this is achieved without costly model preprocessing such as segmentation by parts and medial axis extraction.

In this paper, we propose a novel idea to address the problem of fast computation of enveloping grasp configurations for a multi-fingered hand with 3D polygonal models represented as polygon soups. The proposed method performs a low-level shape matching by wrapping multiple cords around an object in order to quickly isolate promising grasping spots. From these spots, hand palm posture can be computed followed by a standard close-until-contact procedure to find the contact points. Along with the contacts information, the finger kinematics is then used to filter the unstable grasps. Through multiple simulated examples with a twelve degrees-of-freedom anthropomorphic hand, we demonstrate that our method can compute good grasps for objects with complex geometries in a short amount of time. Best of all, this is achieved without complex model preprocessing like segmentation by parts and medial axis extraction.

Robotic exoskeletons can apply forces distributed on the limbs of the subject they are connected to. This offers a great potential in the field of neurorehabilitation, to address the impairment of interjoint coordination in hemiparetic stroke patients. In these patients, the normal flexible joint rotation synergies are replaced by pathological fixed patterns of rotation. In this paper, we investigate how the concept of synergy can be exploited in the control of an upper limb exoskeleton. The long term goal is to develop a device capable of changing the joint synchronization of a patient performing exercises during rehabilitation. The paper presents a controller able of generating joint viscous torques in such a way that constraints on joint velocities can be imposed to the subject without constraining the hand motion. On another hand, the same formalism is used to describe synergies observed on the arm joint motion of subjects realizing pointing tasks. This approach is experimented on a 4 Degrees Of Freedom (DoF) upper arm exoskeleton with subjects performing pointing 3-dimensional tasks. Results exhibit the basic properties of the controller and show its capacity to impose an arbitrary chosen synergy without affecting the hand motion.

This paper presents a novel needle path planning method for biopsy guided by digital breast tomosynthesis (DBT) taking into account breast heterogeneity. First, a multi-resolution optimization approach, guaranteeing that a relevant path is computed, is proposed. Moreover, local breast mechanical parameters required for a heterogeneous finite element simulation, are extracted from the DBT data. The proposed approach computes a 3D local breast glandularity estimation used for Young's modulus determination. This planning method, using a heterogeneous model, reduces the tool positioning error meanly by 75%. In addition, the results show that a breast heterogeneous model has the potential to improve planning accuracy.

This paper presents a new needle path planning method for digital breast tomosynthesis biopsy. Needle insertion planning into deformable tissue for breast biopsy procedure is a challenging task because of the infinite possibilities of insertion points. In addition, the lesion moves from its original position when the radiologist introduces the biopsy needle. The proposed approach couples Rapidly-exploring Random Trees with Finite Element Simulation in order to find an optimal path taking breast deformations into account. Simulation results show that this method reduces the error (i.e. the distance between the needle tip and the lesion) by 80%.

One of the key features of upper limb exoskeletons is their ability to take advantage of the human arm kinematic redundancy in order to modify the subject's joint dynamics without affecting his/her hand motion. This is of particular interest in the field of neurorehabilitation, when an exoskeleton is used to interact with a patient who suffers from joint motions desynchronization, resulting e.g. from brain damage following a stroke. In this paper, we investigate this problem from the robot control point of view. A first general controller is derived which uses viscous force fields in order to generate joint torques counteracting any velocities that are perpendicular to a given set of constraints. In order to minimize energy dissipation, a second controller is proposed that still uses viscous force fields, but in a way that the mechanical power dissipated by the viscous control is null at any time. This controller does not impose any trajectory to the hand and the robot only moves in response to the forces generated by the patient. This approach is experimented on a 4-DOF exoskeleton with a healthy subject. Results exhibit the basic properties of the controller and show its capacity to impose an arbitrary chosen joint constrain for 3-DOF pointing tasks without constraining the hand motion.

The paper deals with computing frictional force-closure grasps of 3D objects problem. The key idea of the presented work is the demonstration that wrenches associated to any three non-aligned contact points of 3D objects form a basis of their corresponding wrench space. This result permits the formulation of a new sufficient force-closure test. Our approach works with general objects, modelled with a set of points, and with any number n of contacts (n ges 4). A quality criterion is also introduced. A corresponding algorithm for computing robust force-closure grasps has been developed. Its efficiency is confirmed by comparing it to the classical convex-hull method [26].

The paper presents a novel strategy that learns to associate a grasp to an unknown object/task. A hybrid approach combining empirical and analytical methods is proposed. The empirical step ensures task-compatibility by learning to identify the object graspable part in accordance with humans choice. The analytical step permits contact points generation guaranteeing the grasp stability. The robotic hand kinematics are also taken into account. The corresponding results are illustrated using GraspIt interface.

The stability of a grasp is characterized by force-closure property, under which arbitrary forces and torques exerted on the grasped object can be balanced by the contact forces applied by the fingers. Existing force-closure tests require considerable computation time. Thus, heuristic approaches are a way to improve performance. In this work, we propose a sufficient but not necessary method to compute force-closure grasps of 3D objects. Our approach works with general objects and with any number n of contacts (nges4).

A grasp is the beginning of any manipulation task. Therefore, an autonomous robot should be able to grasp objects it sees for the first time. It must hold objects appropriately in order to successfully perform the task. This paper considers the problem of grasping unknown objects in the same manner as humans. Based on the idea that the human brain represents objects as volumetric primitives in order to recognize them, the presented algorithm predicts grasp as a function of the object's parts assembly. Beginning with a complete 3D model of the object, a segmentation step decomposes it into single parts. Each single part is fitted with a simple geometric model. A learning step is finally needed in order to find the object component that humans choose to grasp it.

In this paper, we propose a new method for the motion planning problem of rigid object dexterous manipulation with a robotic multi-fingered hand, under quasi-static movement assumption. This method computes both object and finger trajectories as well as the finger relocation sequence. Its specificity is to use a special structuring of the research space that allows to search for paths directly in the particular subspace GSnwhich is the subspace of all the grasps that can be achieved with n grasping fingers. The solving of the dexterous manipulation planning problem is based upon the exploration of this subspace. The proposed approach captures the connectivity of GSnin a graph structure. The answer of the manipulation planning query is then given by searching a path in the computed graph. Simulation experiments were conducted for different dexterous manipulation task examples to validate the proposed method.

This paper addresses the motion planning problem of the dexterous manipulation of 3D rigid objects by a robotic multi-fingered hand. We propose a novel approach based on probabilistic roadmap techniques. Inspired by the theory developed by Alami et al. (1994), Simeon et al. (2003), the planner relies on a topological property that characterizes the existence of solutions in GSn, a specific manifold of the configuration space. This property leads to reduce the problem by structuring the search-space. It allows us to design a manipulation planner that directly captures in a probabilistic roadmap the connectivity of sub-dimensional manifolds of the composite configuration space. The proposed method allows a global planning - both object and fingers trajectories are computed - that can cope with the obstacle presence in the environment. Collisions between different fingers or between object and fingers elsewhere than fingertips are avoided. Force closure constraints are taken into account to ensure the computed paths physical feasibility, under quasi-static motion assumption. First experiments demonstrate the feasibility and the efficiency of the approach

This paper addresses the manipulation planning problem which deals with motion planning for robots manipulating movable objects among static obstacles. We propose a manipulation planner capable of handling continuous domains for modeling both the possible grasps and the stable placements of a single movable object, rather than discrete sets generally assumed by the existing planners. The algorithm relies on a topological property that characterizes the existence of solutions in the subspace of configurations where the robot grasps the object placed at a stable position. This property leads to reduce the problem by structuring the search-space. It allows us to devise a manipulation planner that directly captures in a probabilistic roadmap the connectivity of sub-dimensional manifolds of the composite configuration space. First experiments demonstrate the feasibility and the efficiency of the approach.

An important skill of autonomous robots is the ability to carry out manipulation tasks. The solution to a manipulation problem generally consists in a sequence of elementary paths where an object is moved by a robot or it stays at a stable placement while the robot performs a re-grasping motion. Most existing planners require a finite set of configurations to achieve this task decomposition. We recently proposed an approach to automatically compute such intermediate configurations from continuous sets of stable placements and possible grasps of the movable object. This paper describes an improved algorithm based on this approach. It also presents several complex manipulation problems that illustrate the efficiency of the planner.