# Bringing the Discussion of Minima Sharpness to the Audio Domain: a Filter-Normalised Evaluation for Acoustic Scene Classification

arXiv: https://arxiv.org/abs/2309.16369 · HTML: https://mlsanity.com/paper/2309.16369/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 37 references that are in the map.
- First posted on arXiv 28 Sep 2023.

## Builds on (most cited first, 25 of 25)

- [On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minima](https://mlsanity.com/paper/1609.04836/) (2016)
- [Averaging Weights Leads to Wider Optima and Better Generalization](https://mlsanity.com/paper/1803.05407/) (2018)
- [Visualizing the Loss Landscape of Neural Nets](https://mlsanity.com/paper/1712.09913/) (2017)
- [Optimizing Neural Networks with Kronecker-factored Approximate Curvature](https://mlsanity.com/paper/1503.05671/) (2015)
- [PANNs: Large-Scale Pretrained Audio Neural Networks for Audio Pattern Recognition](https://mlsanity.com/paper/1912.10211/) (2019)
- [Sharp Minima Can Generalize For Deep Nets](https://mlsanity.com/paper/1703.04933/) (2017)
- [SWAD: Domain Generalization by Seeking Flat Minima](https://mlsanity.com/paper/2102.08604/) (2021)
- [Entropy-SGD: Biasing Gradient Descent Into Wide Valleys](https://mlsanity.com/paper/1611.01838/) (2016)
- [PyHessian: Neural Networks Through the Lens of the Hessian](https://mlsanity.com/paper/1912.07145/) (2019)
- [Qualitatively characterizing neural network optimization problems](https://mlsanity.com/paper/1412.6544/) (2014)
- [Robustness via curvature regularization, and vice versa](https://mlsanity.com/paper/1811.09716/) (2018)
- [Towards Theoretically Understanding Why SGD Generalizes Better Than ADAM in Deep Learning](https://mlsanity.com/paper/2010.05627/) (2020)
- [A Modern Look at the Relationship between Sharpness and Generalization](https://mlsanity.com/paper/2302.07011/) (2023)
- [Asymmetric Valleys: Beyond Sharp and Flat Local Minima](https://mlsanity.com/paper/1902.00744/) (2019)
- [Relating Adversarially Robust Generalization to Flat Minima](https://mlsanity.com/paper/2104.04448/) (2021)
- [Acoustic scene classification in DCASE 2020 Challenge: generalization across devices and low complexity solutions](https://mlsanity.com/paper/2005.14623/) (2020)
- [An Encoder-Decoder Based Audio Captioning System With Transfer and Reinforcement Learning](https://mlsanity.com/paper/2108.02752/) (2021)
- [Flatness is a False Friend](https://mlsanity.com/paper/2006.09091/) (2020)
- [Visualizing high-dimensional loss landscapes with Hessian directions](https://mlsanity.com/paper/2208.13219/) (2022)
- [Exploring the Geometry and Topology of Neural Network Loss Landscapes](https://mlsanity.com/paper/2102.00485/) (2021)
- [Device-Robust Acoustic Scene Classification Based on Two-Stage Categorization and Data Augmentation](https://mlsanity.com/paper/2007.08389/) (2020)
- [Why flatness does and does not correlate with generalization for deep neural networks](https://mlsanity.com/paper/2103.06219/) (2021)
- [Exploring loss function topology with cyclical learning rates](https://mlsanity.com/paper/1702.04283/) (2017)
- [Pruning artificial neural networks: a way to find well-generalizing, high-entropy sharp minima](https://mlsanity.com/paper/2004.14765/) (2020)
- [Fairness and underspecification in acoustic scene classification: The case for disaggregated evaluations](https://mlsanity.com/paper/2110.01506/) (2021)

## Built on by (most cited first, 4 of 4)

- [Audio-based Kinship Verification Using Age Domain Conversion](https://mlsanity.com/paper/2410.11120/) (2024)
- [Does the Definition of Difficulty Matter? Scoring Functions and their Role for Curriculum Learning](https://mlsanity.com/paper/2411.00973/) (2024)
- [EnvId: A Metric Learning Approach for Forensic Few-Shot Identification of Unseen Environments](https://mlsanity.com/paper/2405.02119/) (2024)
- [How Class Ontology and Data Scale Affect Audio Transfer Learning](https://mlsanity.com/paper/2603.25476/) (2026)

## Abstract (arXiv)

The correlation between the sharpness of loss minima and generalisation in the context of deep neural networks has been subject to discussion for a long time. Whilst mostly investigated in the context of selected benchmark data sets in the area of computer vision, we explore this aspect for the acoustic scene classification task of the DCASE2020 challenge data. Our analysis is based on two-dimensional filter-normalised visualisations and a derived sharpness measure. Our exploratory analysis shows that sharper minima tend to show better generalisation than flat minima -even more so for out-of-domain data, recorded from previously unseen devices-, thus adding to the dispute about better generalisation capabilities of flat minima. We further find that, in particular, the choice of optimisers is a main driver of the sharpness of minima and we discuss resulting limitations with respect to comparability. Our code, trained model states and loss landscape visualisations are publicly available.