Abstract

International audience; Using 3D computer graphics to emulate watercolor presents a special challenge. Complex stylizations are commonly processed offline, by combining multiple passes in compositing, where art directing is slow and non-intuitive because the stylized result is not immediate. This direct 3D stylization pipeline allows art direction to happen in real time. Using the framework, artists can assign their desired local and global effects directly in the 3D scene, see the stylized results immediately, and intuitively adapt them to fit their stylized vision. The technique can be used and applied in 3D animations, games, VR, visualizations, illustrations, and interactive art.

Abstract

International audience; Advanced Driver Assistance Systems (ADAS) are designed to help drivers improve driving safety. However, automation modifying the way drivers interact with their vehicle, it is important to avoid negative safety impacts. In particular, the change in drivers’ behavior introduced by ADAS in situations they are not designed for, should be carefully examined. We carried out an experiment on a driving simulator to study drivers’ reaction in an obstacle avoidance situation, when using a lateral control assistance system. A detailed analysis of the avoidance maneuver is presented. Results show that assisted and non-assisted drivers equally succeeded in avoiding the obstacle. However, further analyses tend to show an influence of the assistance system on drivers’ first reaction.

Abstract

International audience; With the growth of organic Fruits and Vegetables (FaVs) markets, there is now a trend in marketing research toward studies of non-standardized fruits and vegetables in stores. Yet, because of the decaying nature of FaVs, it is difficult to conduct such studies. A solution is to conduct them within a Virtual Environment (VE) (with virtual FaVs). Therefore, it is of interest to develop an approach to generate a large variety and variability of FaVs, so one can later use them in a VE. In this paper we introduce a pipeline to generate a large variability of FaVs, focusing both on their shape and on their external appearance. Regarding the shape, we use a semi-automated approach. A parametric Skeletal Structure and Generalized Cylinders (GCs) generates their overall shape and metaball-based techniques give them an organic aspect. Regarding their external appearance, we use a particle system approach to simulate their modifications over time. This particle system-based approach decomposes FaVs appearance changes into distinct visual characteristics producing different texture maps that can be combined.

Abstract

International audience; Precise localization in dense urban areas is a challenging task for both mobile mapping and driver assistance systems. This paper proposes a strategy to use road markings as localization landmarks for vision based systems. First step consists in reconstructing a map of road marks. A mobile mapping system equipped with precise georeferencing devices is applied to scan the scene in 3D and to generate an ortho-image of the road surface. A RJMCMC sampler that is coupled with a simulated annealing method is applied to detect occurrences of road marking templates instanced from an extensible database of road mark patterns. The detected objects are reconstructed in 3D using the height information obtained from 3D points. A calibrated camera and a low cost GPS receiver are embedded on a vehicle and used as localization devices. Local bundle adjustment (LBA) is applied to estimate the trajectory of the vehicle. In order to reduce the drift of the trajectory, images are matched with the reconstructed road marks frequently. The matching is initialized by the initial poses that are estimated by LBA and optimized by a MCMC algorithm. The matching provides ground control points that are integrated in the LBA in order to refine the pose parameters. The method is evaluated on a set of images acquired in a real urban area and is compared with a precise ground-truth.