Saved in:
Bibliographic Details
Main Authors: Kosmydel, Jakub, Gajewski, Paweł, Białek, Arkadiusz
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.27105
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913073304961024
author Kosmydel, Jakub
Gajewski, Paweł
Białek, Arkadiusz
author_facet Kosmydel, Jakub
Gajewski, Paweł
Białek, Arkadiusz
contents Analyzing mutual gaze (MG) and joint attention (JA) is critical in developmental psychology but traditionally relies on labor-intensive manual coding. Automating this process in multi-camera laboratory settings is computationally challenging due to complex cross-camera relational dynamics. In this paper, we propose a highly efficient dual-stream Transformer architecture for detecting MG and JA from synchronized dual-camera recordings. Our approach leverages frozen gaze-aware backbones (GazeLLE) to extract rich visual priors, combined with a custom token fusion mechanism to map the spatial and semantic relationships between interacting dyads. Evaluated on an ecologically valid dataset of caregiver-infant interactions, our model exhibits good performance, significantly outperforming both a convolutional baseline and a state-of-the-art multimodal Large Language Model (LLM). By open-sourcing our model and pre-trained weights, we provide behavioral scientists with a scalable tool that can be fine-tuned to diverse laboratory environments, effectively bridging the gap between computational modeling and applied interaction research.
format Preprint
id arxiv_https___arxiv_org_abs_2604_27105
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Automated Detection of Mutual Gaze and Joint Attention in Dual-Camera Settings via Dual-Stream Transformers
Kosmydel, Jakub
Gajewski, Paweł
Białek, Arkadiusz
Computer Vision and Pattern Recognition
Analyzing mutual gaze (MG) and joint attention (JA) is critical in developmental psychology but traditionally relies on labor-intensive manual coding. Automating this process in multi-camera laboratory settings is computationally challenging due to complex cross-camera relational dynamics. In this paper, we propose a highly efficient dual-stream Transformer architecture for detecting MG and JA from synchronized dual-camera recordings. Our approach leverages frozen gaze-aware backbones (GazeLLE) to extract rich visual priors, combined with a custom token fusion mechanism to map the spatial and semantic relationships between interacting dyads. Evaluated on an ecologically valid dataset of caregiver-infant interactions, our model exhibits good performance, significantly outperforming both a convolutional baseline and a state-of-the-art multimodal Large Language Model (LLM). By open-sourcing our model and pre-trained weights, we provide behavioral scientists with a scalable tool that can be fine-tuned to diverse laboratory environments, effectively bridging the gap between computational modeling and applied interaction research.
title Automated Detection of Mutual Gaze and Joint Attention in Dual-Camera Settings via Dual-Stream Transformers
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2604.27105