DinoCompanion: An Attachment-Theory Informed Multimodal Robot for Emotionally Responsive Child-AI Interaction

2025-06-14

AI Summary

Current AI companions lack grounding in developmental psychology, limiting their capacity to provide attachment-based emotional support essential for children aged 3–8. To address deficiencies in developmentally appropriate affective architecture, balanced safety–engagement trade-offs, and attachment-oriented evaluation, this work introduces the first attachment-theory-driven multimodal companion robot framework for children. Our contributions include: (1) CARPO—a risk-calibrated preference optimization objective; (2) AttachSecure-Bench, a novel attachment-competence benchmark (Cohen’s κ = 0.81); and (3) a caregiver–child multimodal dataset comprising 125,000 annotated segments. The system integrates vision–speech–behavior perception, hierarchical memory modeling, and cognitive-uncertainty-weighted risk modeling. Experiments demonstrate a composite attachment-competence score of 57.15% (state-of-the-art), secure-base behavior accuracy of 72.99% (vs. human expert 78.4%), and risk identification accuracy of 69.73%, significantly outperforming GPT-4o and Claude-3.7-Sonnet.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Cognitive Modeling & Cognitive Systems: Affective Computing
• Natural Language Processing: Safety and Robustness

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Security and Privacy: Large-scale security measurements

Abstract

Children's emotional development fundamentally relies on secure attachment relationships, yet current AI companions lack the theoretical foundation to provide developmentally appropriate emotional support. We introduce DinoCompanion, the first attachment-theory-grounded multimodal robot for emotionally responsive child-AI interaction. We address three critical challenges in child-AI systems: the absence of developmentally-informed AI architectures, the need to balance engagement with safety, and the lack of standardized evaluation frameworks for attachment-based capabilities. Our contributions include: (i) a multimodal dataset of 128 caregiver-child dyads containing 125,382 annotated clips with paired preference-risk labels, (ii) CARPO (Child-Aware Risk-calibrated Preference Optimization), a novel training objective that maximizes engagement while applying epistemic-uncertainty-weighted risk penalties, and (iii) AttachSecure-Bench, a comprehensive evaluation benchmark covering ten attachment-centric competencies with strong expert consensus (k{appa}=0.81). DinoCompanion achieves state-of-the-art performance (57.15%), outperforming GPT-4o (50.29%) and Claude-3.7-Sonnet (53.43%), with exceptional secure base behaviors (72.99%, approaching human expert levels of 78.4%) and superior attachment risk detection (69.73%). Ablations validate the critical importance of multimodal fusion, uncertainty-aware risk modeling, and hierarchical memory for coherent, emotionally attuned interactions.

Problem

• Lack of developmentally-informed AI architectures for child emotional support
• Balancing engagement with safety in child-AI interaction systems
• Absence of standardized evaluation frameworks for attachment-based AI capabilities

Innovation

• Multimodal dataset with annotated caregiver-child clips
• CARPO optimizes engagement with risk penalties
• AttachSecure-Bench evaluates attachment competencies