Mind the Context: Continual Learning of Socially Appropriate Robot Actions via Environmental-Social Disentanglement
Rafal Robert Karpinski, Fethiye Irmak Dogan, Nikhil Churamani, Yiming Luo, Maartje M. A. de Graaf, Davide Dell'Anna, Hatice Gunes · N/A · 2026
Framework
N/A
License
N/A
Stars
N/A
Summary
Social robots are expected to operate across diverse environments, where similar arrangements can imply different socially appropriate actions, e.g., starting a conversation may be acceptable in a crowded home but disruptive in an office meeting.
Abstract Summary
Key Points
- Social robots are expected to operate across diverse environments, where similar arrangements can imply different socially appropriate actions, e.g., starting a conversation may be acceptable in a crowded home but disruptive in an office meeting.
- Because such norms and environments cannot all be anticipated in advance, robots require continual learning (CL) to adapt from sequential experience while retaining previously acquired knowledge.
- Prior work has studied CL for generating socially appropriate robot actions, but it has not addressed domain-incremental settings in which the robot incrementally encounters diverse contexts (e.g., living room, meeting room, office, hallway), where both environmental (e.g., whether the space is open or cluttered with furniture) and social cues (e.g., how people or other agents are positioned around the robot) jointly shape the appropriateness of robot actions.
- We address this gap with the Explicit Disentanglement Dual-Branch (EDD) framework
- EDD explicitly separates environmental and social-agent related knowledge and uses replay-based rehearsal to mitigate forgetting while learning the appropriateness of robot actions (e.g., cleaning, serving, starting a conversation) across several indoor domains.
Methodology
We address the problem of enabling a robot to continually learn to assess the social appropriateness of its actions from visual observations across a sequence of distinct indoor locations. Different from prior work on social appropriateness prediction [ 10 , 11 , 15 ] , we formulate the problem in the context of domain incremental learning.
Let $\mathcal{A}=\{a_{1},\dots,a_{K}\}$ denote the set of possible robot actions (e.g. starting a conversation, serving food/drinks, carrying small/large objects).
Experiments
This section introduces our experiments conducted to investigate research questions introduced in Section I , along with evaluation datasets, metrics and implementation details. We evaluate our model on merged OfficeDB and MannersDB+ datasets, which are extensions of MannersDB [ 10 ] for varying environmental contexts and robot embodiments.
Results
Table V reports the overall appropriateness prediction performance of EDD (Ours) compared to baseline methods. EDD achieves the lowest RMSE and the highest PCC and CCC, indicating more accurate and more consistent action appropriateness predictions than the compared baselines. FedLGR yields the highest error and near-zero PCC and CCC, indicating the weakest overall performance among the compared models.
DARE++ achieves the strongest baseline correlations, followed by DUCA.
Sources and Demonstrations
Figures are reproduced from Karpinski et al. See the full paper for experimental details and the project page, when the paper links one, for demonstrations and videos.
Related Papers
A Confidence-Aware Multimodal Fusion Framework for Industrial Human-Robot Collaboration
Xinyu Liu, Qiqi Dong, Boya Jia et al.
A confidence-aware multimodal fusion framework (CAMF) is proposed to realize reliable human intention prediction for industrial human-robot collaboration. This framework fuses four heterogeneous modalities including object 6D pose, gaze, skeletal motion and IMU-based hand motion.
A New Human-Likeness and Comfort Index for Robot Movements Along Prescribed Paths
Rosanna Coccaro, Enrico Ferrentino, Antonio Parziale et al.
As human-robot interaction rapidly spreads in numerous fields, the subject of robot acceptance gains increasing importance.
A Physics-Informed Collision Learning Framework for Collaborative Robot Motion Generation
Chen Cai, Steven Liu
Close-proximity multi-arm manipulation requires collision models that are both geometrically accurate and differentiable enough for real-time optimization.
Aim Short to Reach Far: Your Frozen World Model Can Plan Better Than You Think
Xvyuan Liu, Jianjie Fang, Chen Gao et al.
Planners built on visual world models commonly score each predicted outcome by its distance to the encoded goal image. We show that this target can limit control even with exact dynamics and globally optimal short-horizon search: reaching a goal may require actions that initially move away from it.