See More, Detect Less? Taming Information Leakage in Multi-View Anomaly Detection

• URL: http://arxiv.org/abs/2608.25168v1
• Date: Tue, 25 Aug 2026 21:29:35 GMT
• Title: See More, Detect Less? Taming Information Leakage in Multi-View Anomaly Detection
• Abstract summary: We present GLAD(Global-Local Attention Driven framework), the first framework combining local and global cross-view fusion for anomaly detection.<n>We show that effective multi-view fusion must explicitly restrict the information reaching the decoder.<n>Experiments on Real-IAD and MANTA-Tiny show that GLAD outperforms state-of-the-art methods across sample-, image-, and pixel-level metrics.
• Score: 32.33973839948303
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In multi-view anomaly detection, more cross-view information can actually hurt. When multiple inspection views are naively fused in a reconstruction-based pipeline, normal cues from intact views propagate to the decoder, which faithfully reconstructs anomalous regions, collapsing the reconstruction gap the detector depends on. We call this failure mode \emph{cross-view information leakage} and show that effective multi-view fusion must explicitly restrict the information reaching the decoder. Building on this insight, we present GLAD(Global-Local Attention Driven framework), the first framework combining vision foundation model features with local and global cross-view fusion for multi-view anomaly detection. The Multi-view Merging Attention (MMA) module performs local cross-view fusion at linear complexity with learnable view importance weighting and token-wise gating, letting each view selectively incorporate fine-grained evidence from other views at $\mathcal{O}(N)$ cost. The Object-Guided Attention (OGA) module captures global context by aggregating class tokens from all views into a single object-level representation and broadcasting it back to patch tokens via temperature-scaled sigmoid gating, replacing the original patch representations rather than adding a residual to preserve the reconstruction gap. Experiments on Real-IAD and MANTA-Tiny show that GLAD outperforms state-of-the-art methods across sample-, image-, and pixel-level metrics, confirming that principled information restriction is key to multi-view anomaly reasoning.

Related papers

• GeoMAD: Geometry-Aware Multi-View Anomaly Detection via Deformable Fusion and Distributional Alignment [32.33973839948303]
  Multi-view anomaly detection (MvAD) detects defects by exploiting complementary observations from multiple camera viewpoints.<n>Existing methods typically fall into two extremes: voxel-based fusion provides explicit geometric alignment but requires costly 3D construction and class-specific assumptions.<n>We propose GeoMAD, a unified multi-view, multi-class AD framework that addresses both geometric correspondence deficiency and distributional inconsistency.
  arXiv  Detail & Related papers  (2026-08-27T07:15:15Z)
• IMMoE: Incomplete Multi-View Anomaly Detection via Mixture of View Experts Fusion [13.58401394704324]
  Existing Multi-view Anomaly Detection (MAD) methods assume that all views are completely available and model each view separately.<n>We propose textbfIMMoE, which fuses multi-view information through a multi-view expert network and guides the reconstruction of a single view.<n>Our method achieves state-of-the-art performance on both the RIMAD and Real-IAD datasets.
  arXiv  Detail & Related papers  (2026-07-21T12:19:33Z)
• LEFT: Learnable Fusion of Tri-view Tokens for Unsupervised Time Series Anomaly Detection [53.191369031661885]
  Unsupervised time series anomaly detection aims to build a model for identifying abnormal timestamps without assuming the availability of annotations.<n>We present Learnable Fusion of Tri-view Tokens (LEFT), a unified unsupervised TSAD framework that models anomalies as inconsistencies across complementary representations.<n>Experiments on real-world benchmarks show that LEFT yields the best detection accuracy against SOTA baselines, while achieving a 5x reduction on FLOPs and 8x speed-up for training.
  arXiv  Detail & Related papers  (2026-02-09T13:33:49Z)
• Unsupervised Multi-View Visual Anomaly Detection via Progressive Homography-Guided Alignment [14.782512101141016]
  Unsupervised visual anomaly detection from multi-view images presents a significant challenge.<n>ViewSense-AD (VSAD) learns viewpoint-invariant representations by explicitly modeling geometric consistency across views.<n>Anomaly detection is performed by comparing multi-level features from the diffusion model against a learned memory bank of normal prototypes.
  arXiv  Detail & Related papers  (2025-11-24T05:01:16Z)
• Generalized Deep Multi-view Clustering via Causal Learning with Partially Aligned Cross-view Correspondence [72.41989962665285]
  Multi-view clustering (MVC) aims to explore the common clustering structure across multiple views.<n>However, real-world scenarios often present a challenge as only partial data is consistently aligned across different views.<n>We design a causal multi-view clustering network, termed CauMVC, to tackle this problem.
  arXiv  Detail & Related papers  (2025-09-19T14:31:40Z)
• Multi-View Industrial Anomaly Detection with Epipolar Constrained Cross-View Fusion [15.819291772583393]
  We introduce an epipolar geometry-constrained attention module to guide cross-view fusion.<n>To further enhance the potential of cross-view attention, we propose a pretraining strategy inspired by memory bank-based anomaly detection.<n>We demonstrate that our framework outperforms existing methods on the state-of-the-art multi-view anomaly detection dataset.
  arXiv  Detail & Related papers  (2025-03-14T05:02:54Z)
• Multi-view Aggregation Network for Dichotomous Image Segmentation [76.75904424539543]
  Dichotomous Image (DIS) has recently emerged towards high-precision object segmentation from high-resolution natural images. Existing methods rely on tedious multiple encoder-decoder streams and stages to gradually complete the global localization and local refinement. Inspired by it, we model DIS as a multi-view object perception problem and provide a parsimonious multi-view aggregation network (MVANet) Experiments on the popular DIS-5K dataset show that our MVANet significantly outperforms state-of-the-art methods in both accuracy and speed.
  arXiv  Detail & Related papers  (2024-04-11T03:00:00Z)
• Sketched Multi-view Subspace Learning for Hyperspectral Anomalous Change Detection [12.719327447589345]
  A sketched multi-view subspace learning model is proposed for anomalous change detection. The proposed model preserves major information from the image pairs and improves computational complexity. experiments are conducted on a benchmark hyperspectral remote sensing dataset and a natural hyperspectral dataset.
  arXiv  Detail & Related papers  (2022-10-09T14:08:17Z)
• Self-Supervised Predictive Convolutional Attentive Block for Anomaly Detection [97.93062818228015]
  We propose to integrate the reconstruction-based functionality into a novel self-supervised predictive architectural building block. Our block is equipped with a loss that minimizes the reconstruction error with respect to the masked area in the receptive field. We demonstrate the generality of our block by integrating it into several state-of-the-art frameworks for anomaly detection on image and video.
  arXiv  Detail & Related papers  (2021-11-17T13:30:31Z)
• Multiview Detection with Feature Perspective Transformation [59.34619548026885]
  We propose a novel multiview detection system, MVDet. We take an anchor-free approach to aggregate multiview information by projecting feature maps onto the ground plane. Our entire model is end-to-end learnable and achieves 88.2% MODA on the standard Wildtrack dataset.
  arXiv  Detail & Related papers  (2020-07-14T17:58:30Z)
• Generative Partial Multi-View Clustering [133.36721417531734]
  We propose a generative partial multi-view clustering model, named as GP-MVC, to address the incomplete multi-view problem. First, multi-view encoder networks are trained to learn common low-dimensional representations, followed by a clustering layer to capture the consistent cluster structure across multiple views. Second, view-specific generative adversarial networks are developed to generate the missing data of one view conditioning on the shared representation given by other views.
  arXiv  Detail & Related papers  (2020-03-29T17:48:27Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.