A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss

A novelty detection task involves identifying whether a data point is an outlier, given a training dataset that primarily captures the distribution of inliers. The novel class is usually absent, poorly sampled, or not well defined in the training data. A common technique for anomaly detection at pre...

Full description

Saved in:
Bibliographic Details
Main Authors: Muhammad Asad, Ihsan Ullah, Muhammad Adeel Hafeez, Ganesh Sistu, Michael G. Madden
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11025478/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849334694224592896
author Muhammad Asad
Ihsan Ullah
Muhammad Adeel Hafeez
Ganesh Sistu
Michael G. Madden
author_facet Muhammad Asad
Ihsan Ullah
Muhammad Adeel Hafeez
Ganesh Sistu
Michael G. Madden
author_sort Muhammad Asad
collection DOAJ
description A novelty detection task involves identifying whether a data point is an outlier, given a training dataset that primarily captures the distribution of inliers. The novel class is usually absent, poorly sampled, or not well defined in the training data. A common technique for anomaly detection at present is to use an adversarial network generator to generate an anomaly score for inputs using the reconstruction loss. However, because this technique uses a competitive training process, it can be unreliable, with its performance being inconsistent during each adversarial training step. This inconsistency arises from changes in the network’s ability to detect anomalies. In this paper, we propose a revised framework for generative probabilistic novelty detection. We use a similar adversarial autoencoder-based framework but with a lightweight deep network, a novel training paradigm, and a probabilistic score to compute the reconstruction loss. Our methodology calculates the probability of whether a sample comes from the inlier distribution or not. The proposed approach can be applied to anomaly and outlier detection in images and videos. We present the results on multiple benchmark datasets, including the challenging UCSD Ped2 dataset for video anomaly detection. Our results illustrate that our proposed method learns the inlier classes and differentiates them from the outlier classes effectively, leading to better results than the baseline and state-of-the-art methods in several benchmark datasets.
format Article
id doaj-art-640fe831f6f5414da5d831a38bb6ccfb
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-640fe831f6f5414da5d831a38bb6ccfb2025-08-20T03:45:30ZengIEEEIEEE Access2169-35362025-01-0113985309854110.1109/ACCESS.2025.357708011025478A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction LossMuhammad Asad0https://orcid.org/0000-0002-6398-0320Ihsan Ullah1https://orcid.org/0000-0002-7964-5199Muhammad Adeel Hafeez2https://orcid.org/0000-0002-3593-7448Ganesh Sistu3https://orcid.org/0009-0003-1683-9257Michael G. Madden4https://orcid.org/0000-0002-4443-7285Machine Learning Research Group, School of Computer Science, University of Galway, Galway, IrelandMachine Learning Research Group, School of Computer Science, University of Galway, Galway, IrelandMachine Learning Research Group, School of Computer Science, University of Galway, Galway, IrelandMachine Learning Research Group, School of Computer Science, University of Galway, Galway, IrelandMachine Learning Research Group, School of Computer Science, University of Galway, Galway, IrelandA novelty detection task involves identifying whether a data point is an outlier, given a training dataset that primarily captures the distribution of inliers. The novel class is usually absent, poorly sampled, or not well defined in the training data. A common technique for anomaly detection at present is to use an adversarial network generator to generate an anomaly score for inputs using the reconstruction loss. However, because this technique uses a competitive training process, it can be unreliable, with its performance being inconsistent during each adversarial training step. This inconsistency arises from changes in the network’s ability to detect anomalies. In this paper, we propose a revised framework for generative probabilistic novelty detection. We use a similar adversarial autoencoder-based framework but with a lightweight deep network, a novel training paradigm, and a probabilistic score to compute the reconstruction loss. Our methodology calculates the probability of whether a sample comes from the inlier distribution or not. The proposed approach can be applied to anomaly and outlier detection in images and videos. We present the results on multiple benchmark datasets, including the challenging UCSD Ped2 dataset for video anomaly detection. Our results illustrate that our proposed method learns the inlier classes and differentiates them from the outlier classes effectively, leading to better results than the baseline and state-of-the-art methods in several benchmark datasets.https://ieeexplore.ieee.org/document/11025478/Anomaly detectionadversarial autoencodersreconstruction losslatent spaceprobability distributiongenerative adversarial networks (GANs)
spellingShingle Muhammad Asad
Ihsan Ullah
Muhammad Adeel Hafeez
Ganesh Sistu
Michael G. Madden
A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss
IEEE Access
Anomaly detection
adversarial autoencoders
reconstruction loss
latent space
probability distribution
generative adversarial networks (GANs)
title A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss
title_full A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss
title_fullStr A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss
title_full_unstemmed A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss
title_short A Probabilistic Adversarial Autoencoder for Novelty Detection: Leveraging Lightweight Design and Reconstruction Loss
title_sort probabilistic adversarial autoencoder for novelty detection leveraging lightweight design and reconstruction loss
topic Anomaly detection
adversarial autoencoders
reconstruction loss
latent space
probability distribution
generative adversarial networks (GANs)
url https://ieeexplore.ieee.org/document/11025478/
work_keys_str_mv AT muhammadasad aprobabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT ihsanullah aprobabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT muhammadadeelhafeez aprobabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT ganeshsistu aprobabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT michaelgmadden aprobabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT muhammadasad probabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT ihsanullah probabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT muhammadadeelhafeez probabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT ganeshsistu probabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss
AT michaelgmadden probabilisticadversarialautoencoderfornoveltydetectionleveraginglightweightdesignandreconstructionloss