SASep: Saliency-Aware Structured Separation of Geometry and Feature for Open Set Learning on Point Clouds

2025-06-16

AI Summary

In point cloud open-set recognition (OSR), existing methods rely solely on global features to discriminate known versus unknown classes, neglecting part-level semantic discrepancies—thereby limiting joint performance in known-class classification and unknown-class detection. To address this, we propose a saliency-aware structured disentanglement framework. First, we design a tunable semantic decomposition module that explicitly separates geometric structure features from semantic features. Second, we introduce a geometric synthesis mechanism to generate high-fidelity pseudo-unknown samples. Third, we propose a synthetic-auxiliary boundary separation loss coupled with a contrastive boundary expansion strategy to enable geometric–semantic协同 discrimination. Evaluated on multiple standard point cloud OSR benchmarks, our method significantly outperforms state-of-the-art approaches, simultaneously improving unknown-class recall and known-class accuracy. Extensive experiments validate its strong generalization capability and robustness under diverse distribution shifts.

Technology Category

• Computer Vision: Segmentation
• Search and Optimization: Mixed Discrete/Continuous Search
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• 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

Recent advancements in deep learning have greatly enhanced 3D object recognition, but most models are limited to closed-set scenarios, unable to handle unknown samples in real-world applications. Open-set recognition (OSR) addresses this limitation by enabling models to both classify known classes and identify novel classes. However, current OSR methods rely on global features to differentiate known and unknown classes, treating the entire object uniformly and overlooking the varying semantic importance of its different parts. To address this gap, we propose Salience-Aware Structured Separation (SASep), which includes (i) a tunable semantic decomposition (TSD) module to semantically decompose objects into important and unimportant parts, (ii) a geometric synthesis strategy (GSS) to generate pseudo-unknown objects by combining these unimportant parts, and (iii) a synth-aided margin separation (SMS) module to enhance feature-level separation by expanding the feature distributions between classes. Together, these components improve both geometric and feature representations, enhancing the model's ability to effectively distinguish known and unknown classes. Experimental results show that SASep achieves superior performance in 3D OSR, outperforming existing state-of-the-art methods.

Problem

• Enables models to classify known and identify novel 3D object classes
• Addresses uniform treatment of objects by focusing on semantic part importance
• Improves geometric and feature representation for better open-set recognition

Innovation

• Tunable semantic decomposition for object parts
• Geometric synthesis strategy for pseudo-unknown objects
• Synth-aided margin separation for feature distribution