mlsanityData to 14 Oct 2026

Audio classification with Dilated Convolution with Learnable Spacings

arXiv 2309.13972 ↗First posted 25 Sep 2023cs.SD cs.AI eess.AScited by 1

Not on the map. Why?

Abstract (arXiv)

Dilated convolution with learnable spacings (DCLS) is a recent convolution method in which the positions of the kernel elements are learned throughout training by backpropagation. Its interest has recently been demonstrated in computer vision (ImageNet classification and downstream tasks). Here we show that DCLS is also useful for audio tagging using the AudioSet classification benchmark. We took two state-of-the-art convolutional architectures using depthwise separable convolutions (DSC), ConvNeXt and ConvFormer, and a hybrid one using attention in addition, FastViT, and drop-in replaced all the DSC layers by DCLS ones. This significantly improved the mean average precision (mAP) with the three architectures without increasing the number of parameters and with only a low cost on the throughput. The method code is based on PyTorch and is available at https://github.com/K-H-Ismail/DCLS-Audio

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 26 of its 33 references that are in the map. First posted on arXiv 25 Sep 2023.

Builds on

26 of 33 references.

References · 26

  1. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,475
  2. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
  3. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,045
  4. 2022A ConvNet for the 2020scited by 3,579
  5. 2016WaveNet: A Generative Model for Raw Audiocited by 1,722
  6. 2017Random Erasing Data Augmentationcited by 1,239
  7. 2019Large Batch Optimization for Deep Learning: Training BERT in 76 minutescited by 674
  8. 2021AST: Audio Spectrogram Transformercited by 643
  9. 2019PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognitioncited by 574
  10. 2021VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Textcited by 429
  11. 2021Efficient Training of Audio Transformers with Patchoutcited by 226
  12. 2016FractalNet: Ultra-Deep Neural Networks without Residualscited by 216
14 more
  1. 2023FastViT: A Fast Hybrid Vision Transformer using Structural Reparameterizationcited by 162
  2. 2022Contrastive Audio-Visual Masked Autoencodercited by 146
  3. 2022MetaFormer Baselines for Visioncited by 132
  4. 2022MobileOne: An Improved One millisecond Mobile Backbonecited by 109
  5. 2021PSLA: Improving Audio Tagging with Pretraining, Sampling, Labeling, and Aggregationcited by 82
  6. 2016Very Deep Convolutional Neural Networks for Raw Waveformscited by 68
  7. 2022MAViL: Masked Audio-Video Learnerscited by 59
  8. 2021ERANNs: Efficient Residual Audio Neural Networks for Audio Pattern Recognitioncited by 29
  9. 2021Dilated convolution with learnable spacingscited by 20
  10. 2023Adapting a ConvNeXt model to audio classification on AudioSetcited by 14
  11. 2020Sound Event Detection with Depthwise Separable and Dilated Convolutionscited by 6
  12. 2023Dataset balancing can hurt model performancecited by 5
  13. 2023E-PANNs: Sound Recognition Using Efficient Pre-trained Audio Neural Networkscited by 4
  14. 2023Dilated Convolution with Learnable Spacings: beyond bilinear interpolationcited by 4

Built on by

Cited by year

Citing papers per year: 2024: 1.

1 paper.

  1. 2024Dilated Convolution with Learnable Spacings makes visual models more aligned with humans: a Grad-CAM studycited by 0