Download link for different Gunshot sound : Gunshot_sounds.mp3
Automated classification of Gunshot sounds,like Rifle, Shotgun, Handgun etc, can be used in applications such as our security forces may not even be present within audible range of all gunshots at all times, due to the highly mobile nature of modern combat. However, in these situations, the personnel would still desire real-time knowledge of the origin of unfriendly fire so as to prepare prompt and effective counter-measures.
The most common deep learning based approach for classification of sounds is to convert the audio file to an image, and then use a neural network to process the image. Many of the deep learning networks used for such classification are designed specifically for Spectrograms.
Librosa library is used to convert the gunshots audio into Spectrogram.