This study introduces SENSE, an innovative sensorless external force estimation framework for humanoid robots, built on a model-driven approach. Unlike conventional methods that depend on the simplified Linear Inverted Pendulum (LIP) model and offline data processing, SENSE enables real-time and online estimation. By leveraging centroidal dynamics and integrating angular momentum, our approach accurately infers external forces without the need for Force/Torque (F/T) sensors. The proposed method is validated in simulation using the NAO humanoid robot model within the qiBullet physics engine, under both static standing and realistic dynamic walking conditions. Robustness is evaluated using two external force profiles: constant and time-varying (e.g., sinusoidal). Results show that SENSE provides accurate and stable force estimation, even in challenging scenarios such as foot contact transitions where low-cost sensors like Force-Sensing Resistors (FSRs) typically become unreliable. With its low computational cost and reliance only on onboard sensors, SENSE offers a practical alternative to expensive F/T sensors for medium-sized humanoid platforms. To promote reproducibility and further research, the full implementation is publicly available.

The robustness of object detection algorithms plays a prominent role in real-world applications, especially in uncontrolled environments due to distortions during image acquisition. It has been proven that the performance of object detection methods suffers from in-capture distortions. In this study, we present a performance evaluation framework for the state-of-the-art object detection methods using a dedicated dataset containing images with various distortions at different levels of severity. Furthermore, we propose an original strategy of image distortion generation applied to the MS-COCO dataset that combines some local and global distortions to reach much better performances. We have shown that training using the proposed dataset improves the robustness of object detection by 31.5%. Finally, we provide a custom dataset including natural images distorted from MS-COCO to perform a more reliable evaluation of the robustness against common distortions. The database and the generation source codes of the different distortions are made publicly available1,2.

Bilateral interaction between humans and robots is one of the areas that has attracted much attention in recent years. Automation of human behavior recognition is one of the main steps in achieving this goal. In this regard, in this paper we have designed a process for automatic identification of human gestures. The process consists of two main parts. In the first part, by using the Laban Movement Analysis method, we define a robust descriptor, and in the second part, we determine the robustness of the descriptor using the Dynamic time Warping algorithm. The method proposed in this paper has been tested on four public data-sets namely MSR Action 3D, Florence 3D actions, UTKinect-Action3D and SYSU 3D HUMAN-OBJECT INTERACTION data-sets. Given the results obtained from previous work, the efficiency of the proposed method can be more accurately understood. The results obtained confirm the effectiveness and the performance of our model which outperforms results presented in similar works on action recognition.

Recognizing human actions from video sequences is an active research area in computer vision. This paper describes an effective approach to generate compact and informative representations for action recognition. We design a new action feature descriptor inspired from Laban Movement Analysis method. An efficient preprocessing step based on view invariant human motion is presented. Our descriptor is applied in four known machine learning methods, Random Decision Forest, Multi-Layer Perceptron and Multi-class Support Vector Machines (One-Against-One and One-Against-All). Our proposed approach has been evaluated on two challenging benchmarks of action recognition, Microsoft Research Cambridge-12 (MSRC-12) and MSR-Action3D. We follow the same experimental settings to make a direct comparison between the four classifiers and to show the robustness of our descriptor vector. Experimental results demonstrate that our approach outperforms the state-of-the-art methods.

Interactive robotics is a vast and expanding research field. Interactions must be sufficiently natural, with robots having socially acceptable behavior by Humans, adaptable to user expectations. Thus allowing easy integration in our daily lives in various fields (science, industry, health, etc). Natural interaction during Human-Robot collaborative action needs suitable interaction techniques. In our paper we develop a gesture recognition system for natural and intuitive communication between Human and NAO robot. However recognizing meaningful gesture patterns from whole-body gestures is a complex task. That is why we used the Laban Movement Analysis technique to describe high level gestures for NAO tele-operation. The major contributions of the present work is: (1) an efficient preprocessing step based on view invariant Human motion, (2) a robust descriptor vector based on Laban Movement Analysis technique in order to generate compact and informative representations of Human movement, and (3) a gesture recognition system based on Hidden Markov Model method was applied to teleoperate NAO based on our proper database dedicated to the tele-operation of NAO. Our approach was evaluated with two challenging datasets, Microsoft Research Cambridge-12 (MSRC-12) and UTKinect-Action. Experimental results show that our approach outperforms the state-of-the-art methods.

This work focuses on enabling 3D interaction assistance by adding adaptivity depending on the tasks, objectives, and the general interaction context. We model the context using Conceptual Graphs (CG) based on an ontology. Including CG in our scene manager (Virtools) allows us to add semantic information and to describe the available tools. We handle rules leading to adaptation with a logic programming layer (Prolog+CG) included in the Amine platform. This project is a step towards Intelligent Virtual Environments, which proposes a hybrid solution by adding a separate semantic reasoning to classic environments. The first case study automatically manages few modalities depending on the distance to objects, user movement, available tools and modality risks.

We propose a semantic reasoning engine for context-awareness in classic VR environments. It is currently used to automatically detect user's interests and manage visual enhancements depending on the user's movement.

In this paper, we propose a machine-machine collaboration formalism to support groupware tailorability. Our work is based on the 3C functional model by Ellis that decomposes collaboration between users into communication, coordination and cooperation phases. Through our research, we realized that Web services are powerful distributed components offering the desired tools in order to adapt a groupware system to the real needs of users. Therefore, we use this technology to define a collaboration protocol between machines over the network for implementing tailorability in CSCW systems. A groupware architecture is presented based on the proposed formalism. We argue that a protocol between machines over the internet should be defined in order to exchange common services in real time collaboration.

3D camera tracking is an important issue for many kinds of applications such as augmented reality and robotics navigation. When image sequences are used, tracking needs to be initialized by matching 2D data extracted from images with 3D data representing a priori knowledge of the world. However, this process is difficult and the accuracy of the pose estimation strongly depends on the accuracy of the previous matching step. In this paper, we present two original approaches that achieve a 2D/3D points-based matching in order to initialize the 3D camera tracking. The first one is semi-automatic and requires the intervention of the user. The second one is completely automatic. Both approaches are based on SURF descriptors, which have the advantages of being fast and robust against outliers. A description of these two approaches is given and results obtained from experiments performed on real data are exposed and discussed.

In this paper, we propose a software architecture based on Web services and Software agents for groupware tailorability. Through our literature study, we realize that the property of tailorability has a significant impact on designing collaborative applications. Although online applications in the recent years have been growing exponentially, online collaborative work between users is often supported by software applications that provide static basic functionalities, mostly centered on communication tools (text, audio and video). Hence, adding more sophisticated tools for enriching the collaborative experience, as for example, an integrated environment for task coordination and production, requires manually coding them into the application, which requires a significant effort in order to adapt the system to the real needs of users. In a collaborative context, the application designers are not able to predict all users' needs at design time. To remedy this problem, we propose a tailorable groupware architecture that enables the dynamic integration/composition of services into the collaborative application, gaining both in time and performance. Our work is based on the 3C functional model by Ellis that decomposes collaboration between users into communication, coordination and cooperation spaces. Through our research, we realized that Web services are powerful distributed components offering the desired tools to adapt a groupware to the real needs of users. In this paper, we propose a collaboration protocol based on Web services between machines over the network in order to exchange common services. Based on this protocol, we propose our groupware architecture, U3D, that introduces tailorability in collaboration applications.

Behind its spatial organization, the DNA's molecule conceals a rich functional reality. As a result, the study of the DNA's organization is increasingly attracting scientists. However, researches conducted until now are either local, or global but based on approximations and predictions that make the obtained model less credible. On the other hand, further researches on chromosomes' organization have identified some data and models representing, each one, a part of this organization. Our work aims to find a 3D repesentation that is global and more credible of the DNA's molecule. The molecule will be represented at the chromatin fiber level by means of 3D modeling algorithms. Initially, we have identified the biophysical data to be exploited. Indeed, we will first rely on the biophysical data of the chromatin fiber's structure as the persistence length along with the diameter, the confined volume and the curvature energy in order to build a first simple model of the chromatine.

VR systems need more and more sensors to provide the most efficient 3D interaction in the virtual world. With the multiplication of sensors in these systems, new constraints emerged, especially how to process the huge amount of incoming data and how to estimate a correct interpretation from it. However data processing is often a complex approach. In this paper, we present to you a framework that can be used such an upper-layer on the hardware layer to provide data processing and data fusion to a virtual reality application.

In this paper, we present a research work in the context of a national project, DIGITAL OCEAN for the creation and distribution of multimedia content, where we present tailorable groupware architecture, Oce@nyd, to allow a collaborative collection of multimedia information over the Internet. The Oce@nyd architecture is based on Web services and software agents' integration. This research is, until now, never been exploited in the context of groupware tailorability, hence bringing innovation in the CSCW (Computer Supported Cooperative Work) domain. We apply our research work on DIGITAL OCEAN in order to give users means to collaborate more efficiently, for the purpose of enriching underwater maps with up-to-date multimedia files, as well as means to tailor the services proposed in the system in order to adapt the system to their preferences, and hence enhance collaboration.

Augmented Reality (AR) can provide to a Human Operator (HO) a real help to achieve complex tasks, such as remote control of robots and cooperative teleassistance. Using appropriate augmentations, the HO can interact faster, safer and easier with the remote real world. In this paper, we present an extension of an existing distributed software and network architecture for collaborative teleoperation based on networked human-scaled mixed reality and mobile platform. The first teleoperation system was composed by a VR application and a Web application. However the 2 systems cannot be used together and it is impossible to control a distant robot simultaneously. Our goal is to update the teleoperation system to permit a heterogeneous collaborative teleoperation between the 2 platforms. An important feature of this interface is based on the use of different Virtual Reality platforms and different Mobile platforms to control one or many robots. The first aim with this work is to develop collaborative robot teaching with the use of virtual or real robot.

This paper describes a multimodal tracking system to resolve occlusions in augmented reality applications. The first module of the proposed architecture is composed of a vision based system and allows identification and tracking of visible targets. When targets are partially occluded by scene elements, a second module relieves the vision based module and tracks feature points using a robust algorithm. Finally, a multi-sensors tracking approach is implemented to handle total occlusion of targets and maintains registration even if all markers are not visible. Experimental results and many evaluations have been performed to show the efficiency and robustness of the proposed multimodal approach of tracking and occlusion handling in augmented reality.

In this article, we propose an approach to introduce tailorability in the design of groupware, as the approaches already existing are still ambiguous in putting it forward in CSCW systems. We will present a brief overview on some approaches that deals with tailorability in this field. Then, we will make use of concepts and notions from each in order to integrate them in an innovative, value-added and tailorable architecture. We will discuss the purpose of integrating Internet technologies with software agents while putting it forward in the context of tailorable groupware design.

Human Operators (HO) of telerobotics systems may be able to achieve complex operations with robots. Designing usable and effective Human-Robot Interaction (HRI) is very challenging for system developers and human factors specialists. The search for new metaphors and techniques for HRI adapted to telerobotics systems emerge on the conception of Multimodal HRI (MHRI). MHRI allows to interact naturally and easily with robots due to combination of many devices and an efficient Multimodal Management System (MMS). A system like this should bring a new user's experience in terms of natural interaction, usability, efficiency and flexibility to HRI system. So, a good management of multimodality is very. Moreover, the MMS must be transparent to user in order to be efficient and natural.

In this paper, we present a new groupware architectural model called UD3, which is based on the integration of web services technologies with software agents. The aim is to design a tailorable groupware architecture using the integration of both technologies, thus using properties of each while reinforcing their individual strengths. In fact, agent-oriented technology is claimed to become the next breakthrough in the development and implementation of large-scale complex systems, while Web services are fast emerging technologies for connecting remotely executing programs via well established internet protocols. Web services and agents were originally developed with different standards, thus their integration becomes important in the context of groupware tailorability, giving a totally innovative approach for designing collaborative applications.

In mobile outdoor augmented reality applications, accurate localization is critical to register virtual augmentations over a real scene. Vision-based approaches provide accurate localization estimates but are still too sensitive to outdoor conditions (brightness changes, occlusions, etc.). This drawback can be overcome by adding other types of sensors. In this work, we combine a GPS and an inertial sensor with a camera to provide accurate localization. We will present the calibration process and we will discuss how to quantify the 3D localization accuracy. Experimental results on real data are presented.

Design of complex systems requires multi-user cooperation with virtual reality technologies to provide 3D interactions between users and virtual objects. Recent advances in both virtual reality (VR) systems and computer-supported cooperative work (CSCW) technologies have resulted in a convergence of the appearance of the collaborative virtual environments (CVEs) systems supporting different forms of collaboration and interaction between users. The collaboration in these systems refers to the simultaneous interactions (collaborative interaction) of multiple users on a virtual object in an immersive or semi-immersive virtual environment. In this paper, we propose a method to modeling collaborative 3D interactions that supports group interaction in CVEs. The proposed method is based on group awareness concepts (focus and nimbus) combined astutely with 3D interaction paradigms (navigation, selection and manipulation) to provide us a collaborative 3D interaction model needed to design an adaptive workflow for coordination of multi-user 3D interactions in CVEs.

This paper presents a real-time circular targets tracking approach for camera pose estimation which is based on particle filtering framework. Particle filters are sequential Monte Carlo methods based on point mass (or "particle") representations of probability densities, which can be applied to any state-space model. Their ability to deal with non-linearities and non-Gaussian statistics allows to improve robustness comparing to existing approaches, such as those based on the Kalman filter. We propose to combine a circular fiducials detection algorithm with a particle filter to compute the camera 3D pose parameters. One of the main advantages of our approach comparing to the related camera pose estimation works is its capacity to naturally discard outliers which occur because of either image noise or occlusion. Results from real data in an augmented reality setup are then presented, demonstrating the efficiency and robustness of the proposed method.

This paper gives preliminary results about the utilization of an interaction technique called FOLLOW-ME to fasten the selection task for teleoperation system. The implementation of an interaction between a user and a virtual environment (VE) in virtual reality (VR) may use various techniques. However, in the case of teleoperation, the interaction must be very precise and comfortable for the user. The model associated to the FOLLOW-ME technique splits the virtual environment into three zones in which a specific interaction model is used: a free manipulation zone, a scaled manipulation zone and a precise manipulation zone. Each one of the three zones is characterized by a specific interaction granularity. In the precise manipulation zone, we use the concept of virtual guides in order to assist the user to achieve his task. In this paper, our aim is to show that the FOLLOW-ME technique is well suited for selection in teleoperation tasks. To do this, we have first compared the FOLLOW-ME technique with classical interaction techniques in a virtual environment where different targets are situated at different depth and may move. The preliminary results show that our technique is more efficient than the classical go-go and ray-casting techniques, in a sense that the task is more reproducible and easier to accomplish by the user. In a second stage, we use this result to design selection procedures for the ARITI tele-operation system and show that the use of FOLLOW-ME induces benefits for the user

This paper presents a robust line tracking approach for camera pose estimation which is based on particle filtering framework. Particle filters are sequential Monte Carlo methods based on point mass (or "particle") representations of probability densities, which can be applied to any state-space model. Their ability to deal with non-linearities and non-Gaussian statistics allows to improve robustness comparing to existing approaches, such as those based on the Kalman filter. We propose to use the particle filter to compute the posterior density for the camera 3D motion parameters. The experimental results indicate the effectiveness of our approach and demonstrate its robustness even when dealing with severe occlusion.

One of the key problems in Augmented Reality systems, is the synchronization of real and virtual world. In the context of our research efforts, we need to augment a view of the real world with virtual elements which first requires to determine position and orientation of the point of view and then compute a virtual view according to this location. Augmented reality uses a lot of different sensors in order to estimate camera or operator’s point of view. These sensors could provide samples faster than the mixing of virtual and real informations could be displayed. We expose here a way to take into account samples that are generated during the mixing process by combining augmented reality and virtual reality techniques. This method is using a post-rendering technique involving a texture carried out by a software graphics pipeline to perform this task mainly devoted to optical see-through augmented reality systems. This algorithm can lead to deformations on the final rendered image. We will compute the theoretical errors introduced by this method and confront these results to those obtained with a simulation test-bench which is implementing our proposal.

The major problem with augmented reality (AR) systems using see-through head mounted displays (HMD's) is the end-to-end system delay (or latency). This delay exists because the head tracker, scene generator, and communication links require time to perform their tasks, causing a lag between the measurement of head location and the display of the corresponding virtual objects inside the HMD. One way to eliminate or reduce the latency is to predict future head locations. We propose to use optimal Bayesian algorithms for non-linear/non-Gaussian tracking problems, with a focus on particle filters to predict head motion. Particle filters are sequential Monte Carlo methods based upon point mass (or 'particle') representation of probability densities, which can applied to any state space model, and which generalize the traditional Kalman filtering methods. A SIR particle filter is discussed and compared with the standard EKF through an illustrative example.

One of the key problems in augmented reality systems is registration, that is to say the synchronization of real and virtual world. Augmented reality uses a lot of different sensors in order to estimate camera or operator's point of view. These sensors could provide samples faster than mixing of virtual and real information could be displayed. We expose here a way to fake into account samples that are generated during the mixing process. This method is using a post-rendering technique involving a texture to perform this task. We expose the errors reduction obtained by performing such technique with a simulation test-bench implementing our proposal.

3D free form object recognition is one of the most difficult problems in computer vision. We present a new approach which exploits only one luminance image of a complex object to recognize it in the scene by identifying its appearance in the input image. We construct a photometric (non geometric) projective invariant to perform matching between local regions of the object in the image and those of the model. We propose an original method based on what we called "photometric aspects" to construct a discriminative data base of the 3D object model. We demonstrate the effectiveness of our approach while implementing it with complex free form objects and we present some obtained results.

The described problem is a diagnostic task which must be solved for high-velocity industrial machine tools. Due to environmentally degraded conditions, direct measurements are not possible. Due to the rapidity of the machine, human intervention is not possible in case it causes positional faults. So, an oriented vision solution is proposed. The degraded conditions are vibrations, projections of water and metal chips, dazzling, etc. In this case, there is only one problem that is unable to achieve a diagnostic solution: is it the right piece in the right place? That is why complementary methods are proposed and presented in this paper to answer this question.

This paper explains how to find the attitude of 3D polyhedral object using a virtual model for industrial machining applications. First, the problem is detailed and solutions for each point are studied: camera calibration, image analysis, matching techniques and geometrical methods to find the position and orientation of an object using one video camera. Next, our method is presented, taking into account our particular experimental conditions and reducing the complexity of the problem. So it is possible to find the position of an object in a known reference frame. This result can be useful for diagnostic tasks in industrial machining. The experimental results are presented with both simple objects and complex ones, like a cylinder head. Finally, we present the limits of our method and show how it can be done to solve them.