Quality-Aware Robust Multi-View Clustering for Heterogeneous Observation Noise

2026-02-25

AI Summary

This work addresses the sensitivity of existing deep multi-view clustering methods to heterogeneous noise commonly present in real-world scenarios and their inability to model the continuous variation of noise intensity. To this end, we propose QARMVC, a quality-aware robust multi-view clustering framework that introduces, for the first time, a fine-grained sample-level quality scoring mechanism. Specifically, semantic representations are extracted via the information bottleneck principle, and reconstruction errors are leveraged to quantify noise intensity. Building upon these quality scores, we design a quality-weighted contrastive learning strategy and a high-quality consensus fusion scheme to adaptively suppress noise interference during both feature extraction and view fusion. Extensive experiments demonstrate that QARMVC significantly outperforms state-of-the-art methods across five benchmark datasets, exhibiting exceptional robustness and clustering performance particularly under heterogeneous noise conditions.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Deep multi-view clustering has achieved remarkable progress but remains vulnerable to complex noise in real-world applications. Existing noisy robust methods predominantly rely on a simplified binary assumption, treating data as either perfectly clean or completely corrupted. This overlooks the prevalent existence of heterogeneous observation noise, where contamination intensity varies continuously across data. To bridge this gap, we propose a novel framework termed Quality-Aware Robust Multi-View Clustering (QARMVC). Specifically, QARMVC employs an information bottleneck mechanism to extract intrinsic semantics for view reconstruction. Leveraging the insight that noise disrupts semantic integrity and impedes reconstruction, we utilize the resulting reconstruction discrepancy to precisely quantify fine-grained contamination intensity and derive instance-level quality scores. These scores are integrated into a hierarchical learning strategy: at the feature level, a quality-weighted contrastive objective is designed to adaptively suppress the propagation of noise; at the fusion level, a high-quality global consensus is constructed via quality-weighted aggregation, which is subsequently utilized to align and rectify local views via mutual information maximization. Extensive experiments on five benchmark datasets demonstrate that QARMVC consistently outperforms state-of-the-art baselines, particularly in scenarios with heterogeneous noise intensities.

Problem

• multi-view clustering
• heterogeneous noise
• observation noise
• robustness
• noise intensity

Innovation

• Quality-Aware Learning
• Heterogeneous Observation Noise
• Information Bottleneck
• Multi-View Clustering
• Instance-Level Quality Scoring