BUSSARD: Normalizing Flows for Bijective Universal Scene-Specific Anomalous Relationship Detection

TL;DR

BUSSARD introduces a normalizing flow-based model that effectively detects anomalous relationships in scene graphs by leveraging semantic embeddings and likelihood estimation, outperforming existing methods in accuracy and speed.

Contribution

The paper presents a novel normalizing flow approach for scene graph anomaly detection that is more accurate, faster, and robust to synonyms than previous models.

Findings

Achieves 10% higher AUROC than state-of-the-art.

Runs five times faster than previous models.

Maintains stable performance with synonyms, showing robustness.

Abstract

We propose Bijective Universal Scene-Specific Anomalous Relationship Detection (BUSSARD), a normalizing flow-based model for detecting anomalous relations in scene graphs, generated from images. Our work follows a multimodal approach, embedding object and relationship tokens from scene graphs with a language model to leverage semantic knowledge from the real world. A normalizing flow model is used to learn bijective transformations that map object-relation-object triplets from scene graphs to a simple base distribution (typically Gaussian), allowing anomaly detection through likelihood estimation. We evaluate our approach on the SARD dataset containing office and dining room scenes. Our method achieves around 10% better AUROC results compared to the current state-of-the-art model, while simultaneously being five times faster. Through ablation studies, we demonstrate superior robustness and universality, particularly regarding the use of synonyms, with our model maintaining stable performance while the baseline shows 17.5% deviation. This work demonstrates the strong potential of learning-based methods for relationship anomaly detection in scene graphs. Our code is available at https://github.com/mschween/BUSSARD .