Model-Agnostic Approaches To Handling Noisy Labels When Training Sound Event Classifiers

By Eduardo Fonseca, Frederic Font, Xavier Serra. 2019

10 citations recorded by the source corpus.

Label noise is emerging as a pressing issue in sound event classification. This arises as we move towards larger datasets that are difficult to annotate manually, but it is even more severe if datasets are collected automatically from online repositories, where labels are inferred through automated heuristics applied to the audio content or metadata. While learning from noisy labels has been an active area of research in computer vision, it has received little attention in sound event classification. Most recent computer vision approaches against label noise are relatively complex, requiring complex networks or extra data resources. In this work, we evaluate simple and efficient model-agnostic approaches to handling noisy labels when training sound event classifiers, namely label smoothing regularization, mixup and noise-robust loss functions. The main advantage of these methods is that they can be easily incorporated to existing deep learning pipelines without need for network modifications or extra resources. We report results from experiments conducted with the FSDnoisy18k dataset. We show that these simple methods can be effective in mitigating the effect of label noise, providing up to 2.5% of accuracy boost when incorporated to two different CNNs, while requiring minimal intervention and computational overhead.

Subfields: Artificial Intelligence, Civil and Structural Engineering, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Civil and Structural Engineering

In Signal Processing

Citations in the corpus

FSD50K: An Open Dataset of Human-Labeled Sound Events (2020, methodology): “…directly from waveforms [88, 89]; analysis of label noise mitigation methods leveraging the non-exhaustive labeling of the dev set [32, 42, 90]; multimodal approaches using audio and text information (e.g., using the provided Freesound tags, title, and textual description for the clips)…”
Unsupervised Contrastive Learning of Sound Event Representations (2020, background): “Alternatives to conventional supervised learning include the paradigms of semi-supervised learning [3], few-shot learning [4], learning from noisy labels [5, 6], or self-supervision [7].”
Self-Supervised Learning from Automatically Separated Sound Scenes (2021, background): “For example, label noise issues such as missing, incorrect or inconsistent labels are inherent to this process [1, 2, 3, 4].”
Improving Sound Event Classification by Increasing Shift Invariance in Convolutional Neural Networks (2021, methodology): “We choose mixup because the concept of mixing sounds is an audio-informed operation, and it has been proven useful for SET [2, 28, 3] and other sound event research tasks [29].”
Addressing Missing Labels in Large-Scale Sound Event Recognition Using a Teacher-Student Framework With Loss Masking (2020, background): “Previous work analyses the impact of label noise in these tasks [9, 10], as well as proposes ways to mitigate its negative effect [11, 12, 13, 14].”
Audio tagging with noisy labels and minimal supervision (2019, background): “However, in sound recognition it has received little attention, probably due to the conventional paradigm of learning from small and clean datasets; only a few works directly address the analysis and mitigation of label noise [4, 18, 19].”
A Hybrid Parametric-Deep Learning Approach for Sound Event Localization and Detection (2019, background): “Mixup has been proven successful for sound event classification, even in adverse conditions of corrupted labels [27].”