Evolving from Unknown to Known: Retentive Angular Representation Learning for Incremental Open Set Recognition

2025-09-08

AI Summary

In incremental open-set recognition (IOSR), the continuous arrival of data streams and inaccessibility of historical data cause decision boundary degradation and representation drift. Method: This paper proposes a retention-aware angular representation learning framework. First, it constructs an angular space under an equiangular tight frame to decouple magnitude and direction, enhancing generalization. Second, it introduces a virtual-intrinsic interaction training strategy: virtual boundary classes widen the margin between known and unknown classes, while prototype alignment improves unknown-class characterization. Third, a hierarchical correction mechanism jointly optimizes classification loss, open-set discrimination loss, and angular consistency loss. Results: Experiments on CIFAR-100 and TinyImageNet demonstrate that our method significantly outperforms existing IOSR approaches, establishing—for the first time—a robust benchmark for incremental open-set recognition.

Technology Category

• Computer Vision: Representation Learning for Vision
• Machine Learning: Representation Learning
• Knowledge Representation and Reasoning: Argumentation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Existing open set recognition (OSR) methods are typically designed for static scenarios, where models aim to classify known classes and identify unknown ones within fixed scopes. This deviates from the expectation that the model should incrementally identify newly emerging unknown classes from continuous data streams and acquire corresponding knowledge. In such evolving scenarios, the discriminability of OSR decision boundaries is hard to maintain due to restricted access to former training data, causing severe inter-class confusion. To solve this problem, we propose retentive angular representation learning (RARL) for incremental open set recognition (IOSR). In RARL, unknown representations are encouraged to align around inactive prototypes within an angular space constructed under the equiangular tight frame, thereby mitigating excessive representation drift during knowledge updates. Specifically, we adopt a virtual-intrinsic interactive (VII) training strategy, which compacts known representations by enforcing clear inter-class margins through boundary-proximal virtual classes. Furthermore, a stratified rectification strategy is designed to refine decision boundaries, mitigating representation bias and feature space distortion caused by imbalances between old/new and positive/negative class samples. We conduct thorough evaluations on CIFAR100 and TinyImageNet datasets and establish a new benchmark for IOSR. Experimental results across various task setups demonstrate that the proposed method achieves state-of-the-art performance.

Problem

• Classifying known classes while identifying emerging unknown classes incrementally
• Maintaining discriminative decision boundaries with restricted access to former data
• Mitigating representation drift and feature space distortion during knowledge updates

Innovation

• Retentive angular representation learning for incremental recognition
• Virtual-intrinsic interactive training for compacting known representations
• Stratified rectification strategy to refine decision boundaries