Temp-SCONE: A Novel Out-of-Distribution Detection and Domain Generalization Framework for Wild Data with Temporal Shift

2025-12-04

AI Summary

Existing open-world learning methods (e.g., SCONE) exhibit robustness to covariate and semantic shifts in static settings but struggle with temporal drift in dynamic environments. To address this, we propose Temp-SCONE—the first extension of SCONE incorporating temporal consistency—by introducing an Average Threshold Confidence (ATC)-driven regularization loss. This loss enhances out-of-distribution (OOD) detection capability and cross-temporal prediction stability while preserving energy-based inter-class separation. Temp-SCONE jointly optimizes OOD detection and domain generalization. Empirical evaluation demonstrates that it significantly outperforms SCONE on dynamic datasets, achieving substantial gains in classification accuracy and OOD detection reliability. Moreover, it maintains competitive performance even on non-sequential data, confirming its robustness and versatility.

Technology Category

• Search and Optimization: Learning to Search
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Open-world learning (OWL) requires models that can adapt to evolving environments while reliably detecting out-of-distribution (OOD) inputs. Existing approaches, such as SCONE, achieve robustness to covariate and semantic shifts but assume static environments, leading to degraded performance in dynamic domains. In this paper, we propose Temp-SCONE, a temporally consistent extension of SCONE designed to handle temporal shifts in dynamic environments. Temp-SCONE introduces a confidence-driven regularization loss based on Average Thresholded Confidence (ATC), penalizing instability in predictions across time steps while preserving SCONE's energy-margin separation. Experiments on dynamic datasets demonstrate that Temp-SCONE significantly improves robustness under temporal drift, yielding higher corrupted-data accuracy and more reliable OOD detection compared to SCONE. On distinct datasets without temporal continuity, Temp-SCONE maintains comparable performance, highlighting the importance and limitations of temporal regularization. Our theoretical insights on temporal stability and generalization error further establish Temp-SCONE as a step toward reliable OWL in evolving dynamic environments.

Problem

• Detects out-of-distribution inputs in dynamic environments
• Handles temporal shifts to improve model robustness
• Enhances reliability in open-world learning with temporal regularization

Innovation

• Extends SCONE with temporal consistency regularization
• Uses confidence-driven loss based on Average Thresholded Confidence
• Improves robustness to temporal drift in dynamic environments