One-shot learning for acoustic identification of bird species in non-stationary environments

Acconcjaioco, Michelangelo; Ntalampiras, Stavros

doi:10.1109/icpr48806.2021.9412005

Cited by 8 publications

(4 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additional experimentation on on dealing with very small datasets (<25 examples) would be beneficial to the research community. We hypothesis that Siamese neural networks would result in good model performanace and should be explored across various different species (Chicco, 2021;Acconcjaioco and Ntalampiras, 2021).…”

Section: Discussionmentioning

confidence: 99%

Passive acoustic monitoring of animal populations with transfer learning

Dufourq

Batist

Foquet³

et al. 2022

Ecological Informatics

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Passive acoustic monitoring of animal populations with transfer learning

Dufourq

Batist

Foquet³

et al. 2022

Ecological Informatics

View full text Add to dashboard Cite

“…The explanation methodology we propose in this thesis differs from the ones employed by Zhang, Pardo, andDuan 2019 andAcconcjaioco andNtalampiras 2021 since they use gradient-based methods that tend to focus only on the encoder part of the system. Both these researches limit their exploration to the last convolutional layer of the network not considering the architecture as a whole.…”

Section: Siamese Network Explainabilitymentioning

confidence: 99%

Explaining Siamese Networks in Few-Shot Learning for Audio Data

2022

View full text Add to dashboard Cite

Traditional Machine Learning models are not able to generalize correctly when queried on samples belonging to class distributions that were never seen during training. This is a critical issue, since real world application might need to quickly adapt without the necessity of re-training. To overcome these limitations, few-shot learning frameworks have been proposed and their applicability has been studied widely for computer vision tasks. Siamese Networks learn pairs similarity in form of a metric that can be easily extended on new unseen classes. Unfortunately, the biggest downside of such systems is the lack of explainability. In this thesis we verify the applicability of Siamese Networks in the context of few-shot learning for audio inputs and we propose a novel method to explain their outcomes and assess their robustness. This objective is pursued through a perturbation-based method that quantifies how each input feature contributes to the final outcome by measuring the changes in the mean prediction when such feature is perturbed. We conduct several experiments on two distinct dataset to assess the method ability to explain Siamese Networks outcomes in a C-way one-shot framework. Qualitative and quantitative results demonstrate that our method is able to show common intra-class characteristics and erroneous reliance on silent sections. Classification weaknesses get also uncovered when audio clips are generated from heterogeneous sources and recorded in different environments.

show abstract

“…A matrix assessing the dissimilarities M d can be defined in an analogous way where we aim at minimizing its diagonal. It should be mentioned that the sum of similarity and dissimilarity matrices characterizing the accuracy of a given method is 100%, i.e., M s þ M d ¼ 100 for every element [27].…”

Section: Figures Of Merit and Contrasted Approachmentioning

confidence: 99%

Few-shot learning for modeling cyber physical systems in non-stationary environments

Ntalampiras

Potamitis

2022

Neural Comput & Applic

Self Cite

View full text Add to dashboard Cite

This paper proposes a modeling scheme for cyber physical systems operating in non-stationary, small data environments. Unlike the traditional modeling logic, we introduce the few-shot learning paradigm, the operation of which is based on quantifying both similarities and dissimilarities. As such, we designed a suitable change detection mechanism able to reveal previously unknown operational states, which are incorporated in the dictionary online. We elaborate on spectrograms extracted from high-resolution ultrasound depth sensor timeseries, while the backbone of the proposed method is a Siamese Neural Network. The experimental scenario considers data representing liquid containers for fuel/water when the following five operational states are present: normal, accident, breakdown, sabotage, and cyber-attack. Thorough experiments were carried out assessing every aspect of the present framework and demonstrating its efficacy even when very few samples per class are available. In addition, we propose a probabilistic data selection scheme facilitating one-shot learning. Last but not least, responding to the wide requirement for interpretable AI, we explain the obtained predictions by examining the layer-wise activation maps.

show abstract

One-shot learning for acoustic identification of bird species in non-stationary environments

Cited by 8 publications

References 23 publications

Passive acoustic monitoring of animal populations with transfer learning

Passive acoustic monitoring of animal populations with transfer learning

Explaining Siamese Networks in Few-Shot Learning for Audio Data

Few-shot learning for modeling cyber physical systems in non-stationary environments

Contact Info

Product

Resources

About