This work introduces a flow-matching framework with a linear interpolation path between paired view representations, that replaces diffusion with probability flows between observed and missing views, and provides a formal analysis showing that deterministic ODE flows are inherently better aligned with clustering objectives than diffusion-based stochastic trajectories.
Abstract
Incomplete Multi-View Clustering addresses the problem of clustering multi-modal data when certain views are missing. Recent end-to-end generative approaches leverage diffusion models to recover missing views via stochastic noise-to-data trajectories. While expressive, such mechanisms are not explicitly designed for clustering, as they initialize from cluster-agnostic noise and rely on stochastic denoising dynamics. In this work, we revisit probability path design in end-to-end generative IMVC. We introduce a flow-matching framework with a linear interpolation path between paired view representations, that replaces diffusion with probability flows between observed and missing views. We provide a formal analysis showing that deterministic ODE flows are inherently better aligned with clustering objectives than diffusion-based stochastic trajectories, especially in terms of transport mechanisms that respect class-conditional data distributions and maintain cluster consistency in finite-step regimes. Building upon this insight, we develop an end-to-end IMVC architecture that integrates straight-path flow-matching view completion with cluster-level and entropy-based alignment to enforce cross-view clustering consistency. Extensive experiments on standard IMVC benchmarks demonstrate that the proposed framework establishes new state-of-the-art performance.
Multi-view clustering (MVC) relies on consistency learning to align and fuse multi-view information for building clustering decision boundaries. However, mainstream methods adopt sample-to-sample/distribution/structure similarity smoothing for consistency alignment, which builds upon continuous cluster manifolds with semantic and geometric overlap and reliable paired priors of sample correspondences. They suffer from semantic- and instance-level view-unaligned problems inherent in the real-world data with discrete non-convex structures and unreliable cross-view correspondences. Misled by unaligned noise, they weaken discriminative semantic boundaries via similarity smoothing over false-positive pairs, and degrade discrete non-convex structures by mistakenly bridging distinct cluster manifolds via pseudo-semantic interpolation. Such consistency alignment pursues highly similar representations, distributions and structures, which deviates from the clustering logic of many-to-one partitioning as well as its learning goal for separable semantic boundaries. Guided by the many-to-one principle, we jointly formulate consistency alignment and clustering decision as a novel sample-to-cluster map, termed Multi-view Discrete Optimal Transport. Specifically, MvDOT is instantiated as a cluster-level transport matching framework that first adopts semi-discrete OT to learn a global OT barycenter via aggregating semantic-geometric information from all views, and then transports samples to barycenter-anchored consensus clusters under consistency constraints on semantic assignment and geometric measure. Even with discrete cluster manifolds and unreliable sample correspondences, MvDOT achieves cross-view consistent alignment while preserving inter-cluster semantic boundaries to uncover the underlying cluster structures. Extensive experiments show MvDOT outperforms 10 baselines with higher confidence and stronger robustness in complex MVC tasks.
Yuzhuo Dai, Siwei Wang, Zhibin Dong et al.· IEEE Transactions on Pattern...· 0 citations
Multiview data often suffers from missing views, posing significant challenges for incomplete multiview clustering (IMVC). Existing methods typically perform explicit similarity feature-level imputation using observed views, yet they often fail to capture the shared latent distribution across different views, particularly under high missing rates. In addition, many IMVC models adopt a single-stage framework that jointly optimizes data recovery and clustering, resulting in increased model complexity and computational burden due to full-network backpropagation, thus limiting scalability and training efficiency. To address these issues, we propose a two-stage framework, termed hybrid conditional diffusion with implicit repair for IMVC (HCDIR-IMVC), which simplifies the integration of diffusion models while improving the consistency and robustness of the recovered features. In the first stage, we introduce a unified hybrid conditional diffusion restoration (HC-DR) model that leverages probabilistic learning to dynamically select observed views as conditional inputs for the diffusion process. This enables the model to effectively learn the distribution of missing views and guide their generation. To further enhance cross-view consistency, we incorporate contrastive learning into the diffusion framework. Unlike prior approaches that use separate diffusion models for each view, our method adopts a single hybrid-conditioned model, significantly reducing complexity and improving multiview feature utilization. In the clustering stage, we freeze the imputation model's parameters and remove the decoder to reduce training overhead. A representation and distribution alignment strategy is employed to ensure consistent clustering labels within and across views. Extensive experiments demonstrate that HCDIR-IMVC achieves superior performance in accuracy and computational efficiency, particularly under high missing rates.
Wei-Qing Yan, Kang-Long Liu, Chang Tang et al.· IEEE Transactions on Neural...· 0 citations
Generative models like diffusion and flow matching excel in traffic imputation but suffer from high inference latency, as the learned vector fields typically induce curved generative trajectories requiring multi-step numerical integration. Furthermore, learning the transformation from non-informative priors introduces redundant computational overhead. To address these issues, we propose Low-Rank Prior-Induced Consistency Flow Matching (LOFT) for efficient and effective distribution modeling under highly sparse data. First, we construct a low-rank prior from sparse observations to recover inherent spatiotemporal correlations. Initializing the flow with this informative prior reduces the mapping complexity, allowing the model to focus on fine-grained variations. Second, to enable efficient inference by linearizing generative trajectories, we introduce an uncertainty-aware rectification mechanism. This mechanism resolves the gradient conflict between improving accuracy and trajectory linearization by dynamically arbitrating the optimization trade-off based on the training progress and data uncertainty. Experimental results demonstrate that LOFT surpasses state-of-the-art baselines using an NFE (Number of Function Evaluations) of 2, whereas competing methods typically require 20 to 50 NFE, achieving over a 10× improvement in inference efficiency. The code is available at https://github.com/maoxiaowei97/LOFT.
Xiaowei Mao, Tingrui Wu, Yawen Yang et al.· Proceedings of the 32nd ACM...· 0 citations
Multi-view learning aims to enhance performance by integrating information from multiple sources. While different views offer complementary perspectives, extracting consistent and discriminative representations remains a significant challenge due to discrepancies in representation and presence of noise. Existing methods typically separate the denoising and alignment processes, mapping the denoised heterogeneous views to a shared subspace for alignment. This means that noise may propagate or even amplify during the alignment stage, ultimately leading to suboptimal solutions. To address this issue, we propose center-guided spectral diffusion, which replaces traditional alignment with generative modeling. This method avoids noise propagation in multi-view alignment process and prevents multi-view features be aligned to the noise subspace during fusion process. Specifically, we first performs unconditional diffusion in both the feature space and a low-rank spectral space to learn a stable consensus centre anchor. This anchor is then used to condition a guided generative diffusion process, enabling the model to generate more consistent and realistic sample representations. By combining conditional and unconditional diffusion, the proposed method alleviates the noise amplification problem commonly found in traditional alignment methods. Experimental results show that the proposed method outperforms state-of-the-art methods on several datasets.
Jia-Yuan Wang, Jie Lian, Yong-Quan Shi et al.· Proceedings of the Thirty-Fi...· 0 citations
