2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2019
DOI: 10.1109/iros40897.2019.8968443
|View full text |Cite
|
Sign up to set email alerts
|

A Deep Learning Approach for Multi-View Engagement Estimation of Children in a Child-Robot Joint Attention Task

Abstract: In this work we tackle the problem of child engagement estimation while children freely interact with a robot in their room. We propose a deep-based multi-view solution that takes advantage of recent developments in human pose detection. We extract the child's pose from different RGB-D cameras placed elegantly in the room, fuse the results and feed them to a deep neural network trained for classifying engagement levels. The deep network contains a recurrent layer, in order to exploit the rich temporal informat… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
9
0

Year Published

2020
2020
2024
2024

Publication Types

Select...
6
3

Relationship

0
9

Authors

Journals

citations
Cited by 21 publications
(9 citation statements)
references
References 39 publications
0
9
0
Order By: Relevance
“…Comparatively, the Models investigated in their work can be generalised to a similar application domain as the proposed. Hadfield et al [44] use Deep Learning techniques alongside data captured by a camera to classify engagement levels of children in a classroom setting. This was developed to understand the differences in attention displayed by children described as Typically Developed (TD) and those affected by Autism.…”
Section: Discussionmentioning
confidence: 99%
“…Comparatively, the Models investigated in their work can be generalised to a similar application domain as the proposed. Hadfield et al [44] use Deep Learning techniques alongside data captured by a camera to classify engagement levels of children in a classroom setting. This was developed to understand the differences in attention displayed by children described as Typically Developed (TD) and those affected by Autism.…”
Section: Discussionmentioning
confidence: 99%
“…Rudovic et al [30] proposed the Cul-tureNet model based on the typical ResNet-50 architecture for estimating the engaged or not engaged children of different cultures interacting with NAO robot in robot-assisted therapy for children with Autism Spectrum Condition (ASC). In [31], Hadfield et al proposed a deep learning model consisting of three fully connected layers and a single LSTM layer, while the used features are computed using visual data relative to the position of the child's body parts. e reported results were of almost 80% accuracy, but the limited number (3) of Typical Developed (TD) children can justify the quite low accuracy.…”
Section: Related Workmentioning
confidence: 99%
“…Explicitly learning the temporal dynamics between the features as the mapping to engagement could be performed by deep learning approaches. In Hadfield et al ( 2018 ), a Long Short-Term Memory (LSTM) neural network is employed to classify engagement of children to the task using pose data. LSTM are recurrent neural networks able to capture the different dynamics of time series and they have been shown to be efficient in sequence prediction problems.…”
Section: Perceptionmentioning
confidence: 99%
“…LSTM are recurrent neural networks able to capture the different dynamics of time series and they have been shown to be efficient in sequence prediction problems. These models have been successfully applied to engagement recognition using head movements in Hadfield et al ( 2018 ) and Lala et al ( 2017 ) and facial expression in Dermouche and Pelachaud ( 2019a ). Temporal models such as LSTM and Gated Recurrent Unit (GRU) are compared to static deep leaning approaches as well as logistic regression.…”
Section: Perceptionmentioning
confidence: 99%