FOT-e stands for Field Operational Test- Feature Extraction from Video. The project goal is to enrich the existing datasets, mainly collected in three large European projects (euroFOT, Drive C2X and UDRIVE) and a dataset from FFI-funded Eyes-On-Road. The project aims to enhance the datasets by extracting features from the interior recorded video and adding parametric data describing the driver pose and gaze with the goal to build a world class dataset for interior sensing combined with exterior sensing. The database to be used is the SAFER FOT-database.
The findings of the project will be disseminated with different scopes:
Driver state and engagement
The project will disseminate the result in conference and/or journal papers. The project will be linked to the L3 Pilot where many partners are doing similar proof-of-concepts, although most of them looking at e camera views. This European flagship project will be one platform for disseminating parts of the results.
If the project succeeds on extracting relevant features for passive safety system, this will be a leading dataset that could reach global interest. This would set a new angle to a mature research area where new technologies can give new insights.
The methods in designing and implementing efficient training datasets will be disseminated in relevant arenas, e.g. within AI initiative of Sweden