This blog post presents the dataset for the paper titled “Acoustic Feature Analysis for Wet and Dry Road Surface Classification Using Two-stream CNN”. This research work and the dataset are part of my Ph.D under Universiti Putra Malaysia (UPM) that has been conducted during my secondment at the University of Lincoln.
If you find this dataset useful in your research, please cite:
Siavash Bahrami, Shyamala Doraisamy, Azreen Azman, Nurul Amelina Nasharuddin, and Shigang Yue. 2020. Acoustic Feature Analysis for Wet and Dry Road Surface Classification Using Two-stream CNN. In 2020 4th International Conference on Computer Science and Artificial Intelligence (CSAI 2020). Association for Computing Machinery, New York, NY, USA, 194–200. DOI:https://doi.org/10.1145/3445815.3445847
The acoustic recordings of tyre to road interactions were collected in Millbrook Proving Ground, UK. The 1mile straight and dynamics pad tracks, paved with asphalt, were selected for the recording sessions. Two cars were used for this data collection. Car 1 is a 2012 Toyota Aygo and Car 2 is a 2015 VW Tiguan. The collected data for each car is placed in a directory and named accordingly. In addition, the recordings were done on two different road conditions, wet and dry. Also, each condition was recorded in different illumination levels, once during the day and once during the night.
The compressed RAR file contains 8 dry recordings and 8 wet recordings, structured in directories according to car models, road conditions and illumination levels. The file “1.wave” in the directories are data collected on 1mile straight and “2.wave” are the data collected on dynamics pad.
To download the dataset click on the download button below.
This research has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 691154 STEP2DYNA and No 778602 ULTRACEPT.