AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

GODS: Generalized One-Class Discriminative Subspaces for Anomaly Detection

Jue Wang, Anoop Cherian

ICCV 2019 · Label-efficient and robust learning › OOD and anomaly detection · approach: Unified multi-class anomaly detection · 181 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
generalize a pair of hyperplanes (BODS) to a pair of multi-hyperplane subspaces (GODS) (ablation)BODS · Dash-Cam-Pose (BoW)F1-bar: 0.523 -> 0.553Improved
generalize a pair of hyperplanes (BODS) to a pair of multi-hyperplane subspaces (GODS) (ablation)BODS · Dash-Cam-Pose (TCN)F1-bar: 0.532 -> 0.584Improved
generalize a pair of hyperplanes (BODS) to a pair of multi-hyperplane subspaces (GODS) (ablation)BODS · JHMDBF1: 0.725 -> 0.777Improved
generalize a pair of hyperplanes (BODS) to a pair of multi-hyperplane subspaces (GODS) (ablation)BODS · SonarF1: 0.677 -> 0.762Improved
generalize a pair of hyperplanes (BODS) to a pair of multi-hyperplane subspaces (GODS) (ablation)BODS · Pump (Delft)F1: 0.823 -> 0.892Improved
generalize a pair of hyperplanes (BODS) to a pair of multi-hyperplane subspaces (GODS) (ablation)BODS · UCF-CrimeAUC: 68.26 -> 70.46Improved
generalize a pair of hyperplanes (BODS) to subspaces (GODS) (ablation)BODS · UCF-Crimefalse alarm rate: 2.7 -> 2.1Improved
let the two hyperplanes have independent slopes (BODS) instead of equal slope (Slab-SVM)Slab-SVM · JHMDBF1: 0.643 -> 0.725Improved
use TCN pose embedding instead of bag-of-wordsGODS · Dash-Cam-PoseF1-bar: 0.553 -> 0.584Improved
increase the number of hyperplanes (ablation)GODS · four datasetsF1No clear effect
project data to a low-dimensional subspace then apply SVDD (S-SVDD) instead of GODSS-SVDD · Dash-Cam-Pose (TCN)F1-bar: 0.584 -> 0.464Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy