- The first book of its kind devoted to the emerging field of computer vision in sports
- The definitive reference on this topic, covering ball tracking, player tracking and pose estimation, and the detection of types of specific events and sports
- Presents insights from an international selection of experts in the field
InDeV: In-Depth understanding of accident causation for Vulnerable road users
The InDeV project addresses the second bullet point of the topic MG.3.4. i.e. “… in-depth understanding of road accident causation…”. The main objective of the project is to develop a tool-box for in-depth analysis of accident causation for Vulnerable Road Users (VRU) based on a combined use of accident databases, in-depth accident investigations, surrogate safety indicators, self-reported accidents and naturalistic behavioural data. The tool-box will help to link accident causation factors to VRUs’ accident risk, and provide a solid basis for developing preventive countermeasures and a better input for socio-economic cost calculations of VRU accidents. The proposed approach is to reveal the causational factors by focusing on the process of accident development, thus overcoming the main weakness of the traditional accident data based approach that might find correlations between various factors and accident frequency, but not show the causation chains. It will also employ, to a larger extent, observation of critical traffic events that are similar in process to real accidents, but are relatively more frequent and easier to collect in sufficient quantities. The InDeV project includes the following steps: i) review of methods and identification of the critical sites and road user groups; ii) observation studies at the selected sites; iii) development of technical tools for automated behaviour data collection; iv) analysis of the socio-economical costs; v) compilation of the project results and development of the safety analyst tool-box. The project has a clear focus on VRUs and the course of events in accidents they get injured in. It will provide solid knowledge, help to avoid a skewed view on the problem of VRUs’ safety, and facilitate the proposed tailor-made countermeasures for these groups. Moreover, with the use of surrogate safety indicators, there will be no need to wait for accidents to happen in order to learn how to prevent them from happening.
VAP is co-organizer for ECCV 2014 International Workshop on Soft Biometrics.
Michael Holte won the SCIA 2013 Best PhD Award (best computer vision thesis in Scandinavia 2012-2013).
Congratulations to Michael!
Validation data has just been released. For more info about the challenge, have a look at the poster here here.
For more info see: http://gesture.chalearn.org/
For more info see: http://www.vap.aau.dk/cvsports/