Fast Chirplet Transform to Enhance CNN Machine Listening - Validation on Animal calls and Speech

TL;DR

This paper introduces a Fast Chirplet Transform (FCT) as an efficient bioacoustic representation to pretrain CNNs, significantly reducing training time and improving accuracy in animal call and speech classification tasks.

Contribution

It presents the first algorithm for FCT, demonstrating its computational efficiency and effectiveness in enhancing CNN training for bioacoustic and speech data.

Findings

FCT reduces CNN training time by up to 28%.

FCT improves classification accuracy with a +7.8% MAP score for birds.

FCT accelerates training and enhances performance on speech vowels.

Abstract

The scattering framework offers an optimal hierarchical convolutional decomposition according to its kernels. Convolutional Neural Net (CNN) can be seen as an optimal kernel decomposition, nevertheless it requires large amount of training data to learn its kernels. We propose a trade-off between these two approaches: a Chirplet kernel as an efficient Q constant bioacoustic representation to pretrain CNN. First we motivate Chirplet bioinspired auditory representation. Second we give the first algorithm (and code) of a Fast Chirplet Transform (FCT). Third, we demonstrate the computation efficiency of FCT on large environmental data base: months of Orca recordings, and 1000 Birds species from the LifeClef challenge. Fourth, we validate FCT on the vowels subset of the Speech TIMIT dataset. The results show that FCT accelerates CNN when it pretrains low level layers: it reduces training duration by -28\% for birds classification, and by -26% for vowels classification. Scores are also enhanced by FCT pretraining, with a relative gain of +7.8% of Mean Average Precision on birds, and +2.3\% of vowel accuracy against raw audio CNN. We conclude on perspectives on tonotopic FCT deep machine listening, and inter-species bioacoustic transfer learning to generalise the representation of animal communication systems.