Search papers, labs, and topics across Lattice.
This paper addresses the issue of cross-view information leakage in multi-view anomaly detection, where excessive information from multiple views can degrade detection performance. The authors introduce the GLAD framework, which integrates vision foundation model features with a novel Multi-view Merging Attention (MMA) module and an Object-Guided Attention (OGA) module to effectively manage information flow and enhance detection accuracy. Experimental results demonstrate that GLAD significantly outperforms existing state-of-the-art methods across various metrics, underscoring the importance of controlled information fusion in anomaly detection tasks.
Effective multi-view anomaly detection hinges on restricting information flow, as demonstrated by GLAD's superior performance in managing cross-view data.
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.
