Multi-level Attention Model for Weakly Supervised Audio Classification

Abstract

In this paper, we propose a multi-level attention model to solve the weakly labelled audio classification problem. The objective of audio classification is to predict the presence or absence of audio events in an audio clip. Recently, Google published a large scale weakly labelled dataset called Audio Set, where each audio clip contains only the presence or absence of the audio events, without the onset and offset time of the audio events. Our multi-level attention model is an extension to the previously proposed single-level attention model. It consists of several attention modules applied on intermediate neural network layers. The output of these attention modules are concatenated to a vector followed by a multi-label classifier to make the final prediction of each class. Experiments shown that our model achieves a mean average precision (mAP) of 0.360, outperforms the state-of-the-art single-level attention model of 0.327 and Google baseline of 0.314.

arXiv comment5 pages, 3 figures, Submitted to Eusipco 2018

References in corpus (7)

• Adam: A Method for Stochastic Optimization
• Action Recognition using Visual Attention
• Automatic tagging using deep convolutional neural networks
• Audio Event Detection using Weakly Labeled Data
• DeepOrgan: Multi-level Deep Convolutional Networks for Automated Pancreas Segmentation
• Large-scale weakly supervised audio classification using gated convolutional neural network
• Audio Set classification with attention model: A probabilistic perspective

Cited by in corpus (18)

• A General Survey on Attention Mechanisms in Deep Learning
• PSLA: Improving Audio Tagging with Pretraining, Sampling, Labeling, and Aggregation
• Perceiver: General Perception with Iterative Attention
• Weakly Labelled AudioSet Tagging with Attention Neural Networks
• Audio Retrieval with Natural Language Queries: A Benchmark Study
• Behavioral Use Licensing for Responsible AI
• What Makes Training Multi-Modal Classification Networks Hard?
• Secost: Sequential co-supervision for large scale weakly labeled audio event detection
• Collaborative Deep Learning Across Multiple Data Centers
• SAM-GCNN: A Gated Convolutional Neural Network with Segment-Level Attention Mechanism for Home Activity Monitoring
• An End-to-End Audio Classification System based on Raw Waveforms and Mix-Training Strategy
• Cross-scale Attention Model for Acoustic Event Classification
• Visual Attention for Musical Instrument Recognition
• Advancing Multiple Instance Learning with Attention Modeling for Categorical Speech Emotion Recognition
• Transferring Voice Knowledge for Acoustic Event Detection: An Empirical Study
• A Global-local Attention Framework for Weakly Labelled Audio Tagging
• A Multi-Channel Temporal Attention Convolutional Neural Network Model for Environmental Sound Classification
• Multi-label Sound Event Retrieval Using a Deep Learning-based Siamese Structure with a Pairwise Presence Matrix