Siavash Bahrami is undertaking his PhD at project partner University Putra Malaysia (UPM). Siavash is currently on secondment to the University of Lincoln as an Early Stage Researcher. He is supporting the following Work Packages and Tasks as part of the Ultracept Project which is funded by European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement 778062:

• WP2 – Brain-inspired vision systems for long-range hazard perception

Task 2.3: To develop long range hazard perception methods coping with low light conditions

During my secondment at the University of Lincoln, I am working on the development of a neural network model utilizing acoustic signals. This research will be useful where sound data for road wetness levels estimation would be able to support the development of long-range hazard methods coping with low light conditions.

My preliminary experiment was performed on a dataset of tyre recordings available on “Detecting Road Surface Wetness from Audio with Recurrent Neural Networks” to evaluate the proposed CNN architectures. MFCCs were used as acoustic signal features to classify the wet and dry road for preliminary investigation which resulted in 98.7% accuracy.

In the next phase of my study, I’m going to develop a larger dataset, where recording equipment has been purchased and data collection has just begun on several roads in Lincoln.

I am also looking forward to my next Ultracept secondment with the project partner Visomorphic Technology Ltd.

This Marie Sklodowska-Curie secondment has given me access to facilities and recording equipment needed for compiling the dataset needed for both my PhD studies and the Ultracept project. In addition, the weekly meetings with other members of the project has given me the opportunity to discuss and broaden my knowledge in various related research fields.