Do We Really Need to Learn Representations from In-domain Data for Outlier Detection?

Ding

Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining

et al. 2023

Unsupervised out-of-distribution detection (OOD) seeks to identify out-of-domain data by learning only from unlabeled in-domain data. We present a novel approach for this task -Lift, Map, Detect (LMD) -that leverages recent advancement in diffusion models. Diffusion models are one type of generative models. At their core, they learn an iterative denoising process that gradually maps a noisy image closer to their training manifolds. LMD leverages this intuition for OOD detection. Specifically, LMD lifts an image off its original manifold by corrupting it, and maps it towards the in-domain manifold with a diffusion model. For an out-of-domain image, the mapped image would have a large distance away from its original manifold, and LMD would identify it as OOD accordingly. We show through extensive experiments that LMD achieves competitive performance across a broad variety of datasets.

Section: Algorithm 1 Inpaintmentioning

confidence: 99%

GOOD-D: On Unsupervised Graph Out-Of-Distribution Detection

Ding

Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining

et al. 2023

“…Specifically, algorithm selection is important in unsupervised AD, but limited works [8,67,119] have studied this. We may consider self-supervision [83,95,108] and transfer learning [20] to improve tabular AD as well. Thus, we suggest more focus on large-scale evaluation, task-driven algorithm selection, and data augmentation/transfer for unsupervised AD.…”

Section: Overall Model Performance On Real-world Datasets With Varyin...mentioning

confidence: 99%

ADBench: Anomaly Detection Benchmark

Han¹,

Hu²,

Huang³

et al. 2022

Preprint

Given a long list of anomaly detection algorithms developed in the last few decades, how do they perform with regard to (i) varying levels of supervision, (ii) different types of anomalies, and (iii) noisy and corrupted data? In this work, we answer these key questions by conducting (to our best knowledge) the most comprehensive anomaly detection benchmark with 30 algorithms on 55 benchmark datasets, named ADBench. Our extensive experiments (93,654 in total) identify meaningful insights into the role of supervision and anomaly types, and unlock future directions for researchers in algorithm selection and design. With ADBench, researchers can easily conduct comprehensive and fair evaluations for newly proposed methods on the datasets (including our contributed ones from natural language and computer vision domains) against the existing baselines. To foster accessibility and reproducibility, we fully open-source ADBench and the corresponding results.

“…[26] used the k-nearest neighbors distance between the test input and training set features as an anomaly score. [39] trained a GMM on the normal sample features, which could then identify anomalous samples as low probability areas. PANDA [2] attempts to project the pre-trained features of the normal distribution to another compact feature space employing the DSVDD [40] objective function.…”

Section: Comparisons With State-of-the-artmentioning

confidence: 99%

Fake It Till You Make It: Towards Accurate Near-Distribution Novelty Detection

Mirzaei¹,

Salehi²,

Shahabi³

et al. 2022

Preprint

We aim for image-based novelty detection. Despite considerable progress, existing models either fail or face a dramatic drop under the so-called "near-distribution" setting, where the differences between normal and anomalous samples are subtle. We first demonstrate existing methods experience up to 20% decrease in performance in the near-distribution setting. Next, we propose to exploit a scorebased generative model to produce synthetic near-distribution anomalous data. Our model is then fine-tuned to distinguish such data from the normal samples. We provide a quantitative as well as qualitative evaluation of this strategy, and compare the results with a variety of GAN-based models. Effectiveness of our method for both the near-distribution and standard novelty detection is assessed through extensive experiments on datasets in diverse applications such as medical images, object classification, and quality control. This reveals that our method considerably improves over existing models, and consistently decreases the gap between the near-distribution and standard novelty detection performance. Overall, our method improves the near-distribution novelty detection by 6% and passes the state-of-the-art by 1% to 5% across nine novelty detection benchmarks. The code repository is available at https://github.com/rohban-lab/FITYMI.