We present a biologically-motivated system for the recognition of actions from video sequences. The approach builds on recent work on object recognition based on hierarchical feedforward ...architectures 25, 16, 20 and extends a neurobiological model of motion processing in the visual cortex 10. The system consists of a hierarchy of spatio-temporal feature detectors of increasing complexity: an input sequence is first analyzed by an array of motion- direction sensitive units which, through a hierarchy of processing stages, lead to position-invariant spatio-temporal feature detectors. We experiment with different types of motion-direction sensitive units as well as different system architectures. As in 16, we find that sparse features in intermediate stages outperform dense ones and that using a simple feature selection approach leads to an efficient system that performs better with far fewer features. We test the approach on different publicly available action datasets, in all cases achieving the highest results reported to date.
From clinical knowledge, it has been established that hepatic traumas frequently lead to lethal injuries. In frontal or lateral crash situations, these injuries can be induced by pure deceleration ...effects or blunt trauma due to belt or steering wheel impact. Concerning the liver under frontal decelerations, how could one investigate organ behaviour leading to the injury mechanisms? This work couples experimental organ decelerations measurements (with 19 tests on cadaver trunks) and finite element simulation, provides a first analysis of the liver behaviour within the abdomen. It shows the influence of the liver attachment system that leads to liver trauma and also torsion effects between the two lobes of the liver. Injury mechanisms were evaluated through the four phases of the liver kinematics under frontal impact: (1) postero-anterior translation, (2) compression and sagittal rotation, (3) rotation in the transverse plane and (4) relaxation.
•Achievement goal theory is a new theoretical framework in the driving domain.•Some achievement goals predict sensation seeking and ordinary/aggressive violations.•Performance-approach goal adoption ...positively predict sensation seeking/violations.•Mastery-avoidance goal adoption negatively predict sensation seeking/violations.•Relations between these two goals and violations is mediated by sensation seeking.
Risk taking in driving is a major road safety issue. Understanding the individual psychological differences that may influence risk taking may contribute to better overcome its negative consequences. Recently, four achievement goals were highlighted in the driving domain: striving to drive well or to improve as much as possible (mastery-approach goals), to avoid driving badly or to avoid being a worse driver than before (mastery-avoidance goals), to outperform other drivers (performance-approach goals), and to avoid being a worse driver than other drivers (performance-avoidance goals). The first purpose of the present study was to examine the predictive role of these achievement goals in driving on self-reported drivers’ risk taking (ordinary and aggressive violations) and sensation-seeking. The second purpose of the study was to test the mediating role of sensation seeking between achievement goals adoption in driving and violations. A total of 341 French drivers voluntarily filled out the questionnaires assessing the variables of interest. The main results showed that performance-approach goals adoption was found to positively predict sensation seeking, ordinary and aggressive violations, whereas mastery-avoidance goals adoption was found to negatively predict these variables. The results also highlighted that sensation seeking was a significant mediator of the relationships between the two previous achievement goals (performance-approach and mastery-avoidance goals) and ordinary and aggressive violations. In conclusion, the achievement goal model may now be considered a relevant theoretical framework in the driving literature focusing on risk taking, sensation seeking, and road safety.
In order to investigate injury mechanisms, and to provide directions for road safety system improvements, the HUMOS project has lead to the development of a 3D finite element model of the human body ...in driving position. The model geometry was obtained from a 50th percentile adult male. It includes the description of all compact and trabecular bones, ligaments, tendons, skin, muscles and internal organs. Material properties were based on literature data and specific experiments performed for the project. The validation of the HUMOS model was first achieved on isolated segments and then on the whole model in both frontal and lateral impact situations. HUMOS responses were in good agreement with the experimental data used in the model validation and offers now a wide range of applications from crash simulation, optimization of safety systems, to biomedical and ergonomics.
This paper presents the methodology used to improve knowledge about vulnerable road users' accidents and more specifically pedestrians or cyclists. This work is based on a complete analysis of real ...accidents from three different approaches: in-depth accident investigation, numerical simulation with multibody model and experimental reconstitution with PMHS subject. Accidents chosen from an in-depth multidisciplinary investigation are numerically modelled using a multibody software. Then, a parametric study focused for instance on car velocity and victim position at impact is performed in order to find the best correlations with all indications produced by the in-depth analysis. Finally, the retained configuration close to the presumed real accident conditions is reproduced experimentally by a crash test using cadavers. All results are finally compared in order to validate the real accident reconstruction. This methodology is applied on two real accidents involving one pedestrian or one cyclist.
•51 French company vehicles have been equipped with EDR.•More than 100000km and 300 incidents have been recorded.•Around 70% incidents are in lateral, 25% in longitudinal.•The incidents collection ...improves the road safety diagnosis.•We progress towards understanding incident and accident mechanisms.
Because motor vehicle crashes have decreased during the last decade in many countries in the world and are more diffuse, local authorities have difficulties to define road safety policies. An experiment with 51 cars of public fleets equipped with a specific Event Data Recorder was carried out in France during one year. The purposes of this research were to evaluate if incident data (critical driving situations) help to understand crashes, and to explore a new way for road infrastructure safety diagnosis. The analysis of 339 genuine incidents and 1237 simple events recorded illustrates the potentiality of such an experiment and provides: some insights about conditions in which incidents occur, a general overview of their distribution according to different road layouts, as well as information on the different levels of accelerations reached. It can be noticed that there is an overrepresentation of incidents in right curves compared to left curves. The simple events involving mostly the infrastructure could be used to detect road defects. Genuine incidents where the vehicle is subjected to important dynamic demands, related to potentially unsafe driving situations, can be used to improve knowledge of the motor vehicle crashes thanks to incident mechanisms analysis.