Abstract
Open-set recognition (OSR) is a challenging machine learning problem that appears when classifiers are faced with test instances from classes not seen during training. It can be summarized as the problem of correctly identifying instances from a known class (seen during training) while rejecting any unknown or unwanted samples (those belonging to unseen classes). Another problem arising in practical scenarios is few-shot learning (FSL), which appears when there is no availability of a large number of positive samples for training a recognition system. Taking these two limitations into account, a new dataset for OSR and FSL for audio data was recently released to promote research on solutions aimed at addressing both limitations. This paper proposes an audio OSR/FSL system divided into three steps: a high-level audio representation, feature embedding using two different autoencoder architectures and a multi-layer perceptron (MLP) trained on latent space representations to detect known classes and reject unwanted ones. An extensive set of experiments is carried out considering multiple combinations of openness factors (OSR condition) and number of shots (FSL condition), showing the validity of the proposed approach and confirming superior performance with respect to a baseline system based on transfer learning.
Highlights
Machine listening is the branch of artificial intelligence that aims to create intelligent systems that are capable of extracting relevant information from audio data
The 2019 edition incorporated an open-set recognition (OSR) task within the scope of acoustic scene classification (ASC), where the idea was to classify an audio clip to a known scene type or to reject it when it belonged to an unknown scene
This paper proposes a novel deep learning approach to tackle OSR and Few-shot learning (FSL) problems within an Acoustic event classification (AEC) context, based on a combined two-stage method
Summary
Machine listening is the branch of artificial intelligence that aims to create intelligent systems that are capable of extracting relevant information from audio data. The increase in research proposals related to these areas is motivated by the number of applications that can benefit from automation systems incorporating audio-based solutions, such as home assistants or autonomous driving. This interest is evidenced by the multiple editions of the successful international DCASE challenge (Detection and Classification of Acoustic Scenes and Events). The use of autoencoders represents a solution based on embedding learning, that is, a model that is capable of discovering important structure within the input data by forcing a reduction of dimensionality. For a complete review of FSL approaches, the reader is referred to [23]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
