The vehicle whistle database was recorded by Communication Acoustics Group in Institute of Acoustics, Chinese Academy of Sciences. The aim of this database was to allow for evaluation of whistling vehicle localization using real recordings. This database was divided into 5 folders corresponding to 5 tasks based on the number of vehicles and their status of movement, which are:
Task 1: Detection and localization of one static whistling vehicle.
Task 2: Detection and localization of two static whistling vehicles.
Task 3: Detection and tracking of one moving whistling vehicle.
Task 4: Detection and tracking of one static whistling vehicle and one moving whistling vehicles.
Task 5: Training and testing deep learning-based vehicle whistle detection models.