Streaming Network for Continual Learning of Object Relocations under Household Context Drifts

2024-11-08

arXiv.org

AI Summary

To address non-i.i.d. data arising from heterogeneous user behavior patterns and catastrophic forgetting in long-term deployment of domestic service robots, this paper proposes STREAK, a streaming continual learning framework. STREAK innovatively integrates an online spatiotemporal graph neural network, an adaptive knowledge-preserving regularization (an EWC variant), and a lightweight selective experience replay mechanism to enable continual, incremental updating of object relocalization knowledge under dynamic household conditions. Evaluated on real-world trajectory prediction across 50+ days in multiple homes, STREAK achieves a 23% improvement in prediction accuracy and a 67% reduction in forgetting rate over baseline methods, supporting long-term online deployment without full retraining. Its core contribution is the first spatiotemporally aware continual learning architecture specifically designed for home environments, effectively mitigating both distributional shift and knowledge forgetting.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Intelligent Robots: Learning & Optimization for ROB
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

In real-world settings, robots are expected to assist humans across diverse tasks and still continuously adapt to dynamic changes over time. For example, in domestic environments, robots can proactively help users by fetching needed objects based on learned routines, which they infer by observing how objects move over time. However, data from these interactions are inherently non-independent and non-identically distributed (non-i.i.d.), e.g., a robot assisting multiple users may encounter varying data distributions as individuals follow distinct habits. This creates a challenge: integrating new knowledge without catastrophic forgetting. To address this, we propose STREAK (Spatio Temporal RElocation with Adaptive Knowledge retention), a continual learning framework for real-world robotic learning. It leverages a streaming graph neural network with regularization and rehearsal techniques to mitigate context drifts while retaining past knowledge. Our method is time- and memory-efficient, enabling long-term learning without retraining on all past data, which becomes infeasible as data grows in real-world interactions. We evaluate STREAK on the task of incrementally predicting human routines over 50+ days across different households. Results show that it effectively prevents catastrophic forgetting while maintaining generalization, making it a scalable solution for long-term human-robot interactions.

Problem

• Robots adapt to dynamic changes in household object relocations.
• Learning non-i.i.d. data without catastrophic forgetting is challenging.
• Streaming framework enables efficient long-term knowledge retention.

Innovation

• Streaming graph neural network for continual learning
• Regularization and rehearsal to retain knowledge
• Time- and memory-efficient long-term learning