Antoine Seilles

A constrained Extended Kalman Filter for dynamically consistent inverse kinematics and inertial parameters identification

This paper presents a method for the real-time determination of joint angles, velocities, accelerations and joint torques of a human. The proposed method is based on a constrained Extended Kalman Filter that combines stereophotogrammetric and dynamometric data. In addition to the joint variables, subject-specific segment lengths and inertial parameters are identified. Constraints are added to the filter, by restricting the optimal Kalman gain, in order to obtain physically consistent parameters. An optimal tuning procedure of the filters gains and a sensitivity analysis is presented. The method is validated in the plane on four human subjects and shows very good tracking of skin markers with a RMS difference lower than 15 mm. External ground reaction forces and resultant moment are also accurately estimated with an RMS difference below 3 N and 6 N.m, respectively.

Game Design for All: The Example of Hammer and Planks

The last years have seen a growing interest in the Serious Games topic - and in particular in Games for Health - from both scientific and industrial communities. However not only is the effectiveness of this kind of game not yet demonstrated but the distribution and adoption of these games by the mainstream market is still very low. In this paper we present a game for hemiplegic rehabilitation called “Hammer and Planks”. The game was developed with the adoption by the general public in mind and has shown interesting results during a first experimentation at a game exhibition.

Music Games: Potential Application and Considerations for Rhythmic Training

Rhythmic skills are natural and widespread in the general population. The majority can track the beat of music and move along with it. These abilities are meaningful from a cognitive standpoint given their tight links with prominent motor and cognitive functions such as language and memory. When rhythmic skills are challenged by brain damage or neurodevelopmental disorders, remediation strategies based on rhythm can be considered. For example, rhythmic training can be used to improve motor performance (e.g., gait) as well as cognitive and language skills. Here, we review the games readily available in the market and assess whether they are well-suited for rhythmic training. Games that train rhythm skills may serve as useful tools for retraining motor and cognitive functions in patients with motor or neurodevelopmental disorders (e.g., Parkinson’s disease, dyslexia, or ADHD). Our criteria were the peripheral used to capture and record the response, the type of response and the output measure. None of the existing games provides sufficient temporal precision in stimulus presentation and/or data acquisition. In addition, games do not train selectively rhythmic skills. Hence, the available music games, in their present form, are not satisfying for training rhythmic skills. Yet, some features such as the device used, the interface or the game scenario provide good indications for devising efficient training protocols. Guidelines are provided for devising serious music games targeting rhythmic training in the future.

Rhythm Workers: A music-based serious game for training rhythm skills

Rhythm perception and production can be disrupted by neurological or neurodevelopmental disorders (e.g., Parkinson’s disease, dyslexia). Rhythm deficits are associated with poor performance in language, attention, and working memory tasks. Re-training rhythmic skills may thus provide a promising avenue for improving these associated cognitive functions. To this end, here we present a new protocol for selective training of rhythmic skills implemented in a tablet serious game called Rhythm Workers. Experiment 1 served to select 54 musical excerpts based on the tapping performance of 18 non-musicians who moved to the beat of music. The excerpts were sorted in terms of the difficulty of tracking their beat, and assigned to different difficulty levels in the game. In Experiment 2, the training protocol was devised and tested in a proof-of-concept study, including two versions of the game. One version (tapping version) required a synchronized motor response (via tapping), while the other (perception version) asked for a perceptual judgment. Ten participants were trained with one version and 10 with the other version of Rhythm Workers, for 2 weeks. A control group (n = 10) did not receive any training. Participants in the experimental groups showed high compliance and motivation in playing the game. The effect of the training on rhythm skills yielded encouraging results with both versions of the game. Rhythm Workers thus appears to be a motivating and potentially efficient way to train rhythmic abilities in healthy young adults, with possible applications for (re)training these skills in individuals with rhythm disorders.

Visual analysis of body movement in serious games for healthcare

The advancement of motion sensing input devices has enabled the collection of multivariate time-series body movement data. Analyzing such type of data is challenging due to the large amount of data and the task of mining for interesting temporal movement patterns. To address this problem, we propose an interface to visualize and analyze body movement data. This visualization enables users to navigate and explore the evolution of movement over time for different movement areas. We also propose a clustering method based on hierarchical clustering to group similar movement patterns. The proposed visualization is illustrated with a case study which demonstrates the ability of the interface to analyze body movements.

Automatic estimate of back anatomical landmarks and 3D spine curve from a Kinect sensor

This study aims to develop and evaluate a new method for the automatic extraction and estimate of back anatomical landmark positions and of 3D spine curve from Kinect sensor data. The proposed method allows to robustly reconstruct different indexes of back deformity used in the evaluation of scoliosis. The algorithm input data are the depth map and its corresponding curvature map. From these, regions-of-interest are automatically created and anatomical landmark positions are estimated by finding common patterns between subjects. The results showed that the proposed method can successfully estimate the anatomical landmark positions, as well as the 3D spine curve (average RMS error of 8 mm and 3 mm). The simplicity and generalisation abilities of the proposed method allow to pave the way of future diagnosis solutions for in-home or for small size practice use.

Tabu search for human pose recognition

The use of computer vision techniques to build hands-free input devices has long been a topic of interest to researchers in the field of natural interaction. In recent years Microsoft’s Kinect has brought these technologies to the layman, but the most commonly used libraries for Kinect human pose recognition are closed-source. There is not yet an accepted, effective open-source alternative upon which highly specific applications can be based. We propose a novel technique for extracting the appendage configurations of users from the Kinect camera’s depth feed, based on stochastic local search techniques rather than per-pixel classification.

Stakeholder detection for online debates

This paper presents a work in progress about stakeholder detection for online debates. We propose an approach based on classical community detection methods applied to semantic social networks representation. We defend that new web2.0 tools should assist users to define semantic relations between users, groups and roles based on social interaction analysis. The main goal is to provide new mecanisms for moderation decreasing misunderstandings and highlighting unexpected behaviors.