General-purpose Tagging of Freesound Audio with AudioSet Labels: Task Description, Dataset, and Baseline

Abstract

This paper describes Task 2 of the DCASE 2018 Challenge, titled "General-purpose audio tagging of Freesound content with AudioSet labels". This task was hosted on the Kaggle platform as "Freesound General-Purpose Audio Tagging Challenge". The goal of the task is to build an audio tagging system that can recognize the category of an audio clip from a subset of 41 diverse categories drawn from the AudioSet Ontology. We present the task, the dataset prepared for the competition, and a baseline system.

Camera ready for DCASE Workshop 2018

References in corpus (1)

• SoundNet: Learning Sound Representations from Unlabeled Video

Cited by in corpus (24)

• Polyphonic Sound Event Detection and Localization using a Two-Stage Strategy
• Cough Against COVID: Evidence of COVID-19 Signature in Cough Sounds
• Weakly Labelled AudioSet Tagging with Attention Neural Networks
• PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition
• Project Achoo: A Practical Model and Application for COVID-19 Detection from Recordings of Breath, Voice, and Cough
• General audio tagging with ensembling convolutional neural network and statistical features
• DCASE 2018 Challenge Surrey Cross-Task convolutional neural network baseline
• Addressing Missing Labels in Large-Scale Sound Event Recognition Using a Teacher-Student Framework With Loss Masking
• Anomalous Sound Detection with Machine Learning: A Systematic Review
• Sound Event Detection with Sequentially Labelled Data Based on Connectionist Temporal Classification and Unsupervised Clustering
• Adversarial Attacks in Sound Event Classification
• Multichannel-based learning for audio object extraction
• Combining High-Level Features of Raw Audio Waves and Mel-Spectrograms for Audio Tagging
• Multi-Task Learning for Interpretable Weakly Labelled Sound Event Detection
• Listen to What You Want: Neural Network-based Universal Sound Selector
• Diversity-Robust Acoustic Feature Signatures Based on Multiscale Fractal Dimension for Similarity Search of Environmental Sounds
• Attribution in Scale and Space
• Interpretable Super-Resolution via a Learned Time-Series Representation
• Impact of data-splits on generalization: Identifying COVID-19 from cough and context
• Facilitating the Manual Annotation of Sounds When Using Large Taxonomies
• Client Adaptation improves Federated Learning with Simulated Non-IID Clients
• Semi Supervised Learning For Few-shot Audio Classification By Episodic Triplet Mining
• ARCA23K: An audio dataset for investigating open-set label noise
• Few-shot learning of new sound classes for target sound extraction