Acoustic Scene Classification
37 papers with code • 5 benchmarks • 10 datasets
The goal of acoustic scene classification is to classify a test recording into one of the provided predefined classes that characterizes the environment in which it was recorded.
Source: DCASE 2019 Source: DCASE 2018
Benchmarks
These leaderboards are used to track progress in Acoustic Scene Classification
| Trend | Dataset | Best Model | Paper | Code | Compare |
|---|---|---|---|---|---|
|
|||||
|
|||||
|
|||||
|
|||||
|
Datasets
DCASE 2016
MTG-Jamendo
TAU Urban Acoustic Scenes 2019
TUT Acoustic Scenes 2017
DCASE 2013
DCASE 2019 Mobile
LITIS Rouen
CochlScene
TUT Sound Events 2018
Latest papers
A Passive Similarity based CNN Filter Pruning for Efficient Acoustic Scene Classification
Arshdeep-Singh-Boparai/passive-filter-pruning-Interspeech22
•
•
We propose a passive filter pruning framework, where a few convolutional filters from the CNNs are eliminated to yield compressed CNNs.
Acoustic scene classification using auditory datasets
The approach used not only challenges some of the fundamental mathematical techniques used so far in early experiments of the same trend but also introduces new scopes and new horizons for interesting results.
Towards Audio Domain Adaptation for Acoustic Scene Classification using Disentanglement Learning
jakobabesser/icassp_2022_disentanglementasc
•
•
The deployment of machine listening algorithms in real-life applications is often impeded by a domain shift caused for instance by different microphone characteristics.
A Variational Bayesian Approach to Learning Latent Variables for Acoustic Knowledge Transfer
mihawkhu/asc_knowledge_transfer
•
•
We propose a variational Bayesian (VB) approach to learning distributions of latent variables in deep neural network (DNN) models for cross-domain knowledge transfer, to address acoustic mismatches between training and testing conditions.
Efficient Training of Audio Transformers with Patchout
kkoutini/passt
•
•
However, one of the main shortcomings of transformer models, compared to the well-established CNNs, is the computational complexity.
Low-complexity acoustic scene classification for multi-device audio: analysis of DCASE 2021 Challenge systems
The most used techniques among the submissions were residual networks and weight quantization, with the top systems reaching over 70% accuracy, and log loss under 0. 8.
Receptive Field Regularization Techniques for Audio Classification and Tagging with Deep Convolutional Neural Networks
As state-of-the-art CNN architectures-in computer vision and other domains-tend to go deeper in terms of number of layers, their RF size increases and therefore they degrade in performance in several audio classification and tagging tasks.
Spectrum Correction: Acoustic Scene Classification with Mismatched Recording Devices
SRPOL-AUI/spectrum-correction
•
•
This method works for both time and frequency domain representations of audio recordings.
Low-Complexity Models for Acoustic Scene Classification Based on Receptive Field Regularization and Frequency Damping
Deep Neural Networks are known to be very demanding in terms of computing and memory requirements.
A Two-Stage Approach to Device-Robust Acoustic Scene Classification
To improve device robustness, a highly desirable key feature of a competitive data-driven acoustic scene classification (ASC) system, a novel two-stage system based on fully convolutional neural networks (CNNs) is proposed.
