Group-CLIP Uncertainty Modeling for Group Re-Identification

2025-02-10

AI Summary

Group ReID aims to match pedestrian groups across camera views, but faces fundamental challenges due to dynamic variations in group size and spatial configuration—conditions under which existing deterministic models exhibit poor generalization to unseen group structures. To address this, we propose the first CLIP-based framework for Group ReID, introducing a novel uncertainty-aware modeling paradigm. Specifically, we design a Bernoulli-distribution-driven member variation simulation module to explicitly model stochastic member occlusion or absence; an identity-aware uncertain text generator that adaptively describes group composition and topology; and a group-relation-guided individual feature enhancement encoder. These components are jointly optimized via cross-modal contrastive learning to align visual group representations with generated textual descriptions. Extensive experiments demonstrate significant improvements over state-of-the-art methods on multiple benchmarks, particularly under unknown group configurations—achieving superior robustness, generalization, and identification accuracy.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Group Re-Identification (Group ReID) aims matching groups of pedestrians across non-overlapping cameras. Unlike single-person ReID, Group ReID focuses more on the changes in group structure, emphasizing the number of members and their spatial arrangement. However, most methods rely on certainty-based models, which consider only the specific group structures in the group images, often failing to match unseen group configurations. To this end, we propose a novel Group-CLIP UncertaintyModeling (GCUM) approach that adapts group text descriptions to undetermined accommodate member and layout variations. Specifically, we design a Member Variant Simulation (MVS)module that simulates member exclusions using a Bernoulli distribution and a Group Layout Adaptation (GLA) module that generates uncertain group text descriptions with identity-specific tokens. In addition, we design a Group RelationshipConstruction Encoder (GRCE) that uses group features to refine individual features, and employ cross-modal contrastive loss to obtain generalizable knowledge from group text descriptions. It is worth noting that we are the first to employ CLIP to GroupReID, and extensive experiments show that GCUM significantly outperforms state-of-the-art Group ReID methods.

Problem

• Group Re-Identification across cameras
• Addressing unseen group configurations
• Uncertainty modeling for member and layout variations

Innovation

• Group-CLIP Uncertainty Modeling approach
• Member Variant Simulation module
• Group Layout Adaptation module