Trajectory representation learning (TRL) is an intermediate step in handling trajectory data to realize various downstream machine-learning tasks. While most previous TRL research focuses on modeling structured movements in large-scale urban spaces (e.g., cars or pedestrians on streets), this paper focuses on a more challenging scenario of modeling free movement in small-scale social spaces (e.g., children playing in a schoolyard). We present a TRL model, SiamCircle, to process raw trajectories without additional feature extraction to prevent information loss. SiamCircle adopts a Siamese network with Circle Loss to learn trajectory embeddings. Furthermore, SiamCircle employs a data augmentation process to enable self-supervised learning and enrich the input data to address the limited access to high-quality data and ground truth. We evaluate the performance of SiamCircle in downstream tasks using trajectory ranking and clustering performance via seven evaluation metrics collectively. Using an ablation study, we explored the impact of different loss functions on the model’s performance. Accordingly, we selected a 2-D convolutional design with Circle Loss as the best-performing model. In a comparative study, we compared our model against three other baselines. We observed up to 19% improvements in trajectory ranking tasks and achieved the highest average rank in supervised clustering tasks.

Background: Outdoor social participation in the school playground is crucial for children's socio-emotional and cognitive development. Yet, many children with disabilities in mainstream educational settings are not socially included within their peer group. We examined whether loose-parts-play (LPP), a common and cost-effective intervention that changes the playground play environment to enhance child-led free play, can promote social participation for children with and without disabilities. Method: Forty-two primary school children, out of whom three had hearing loss or autism, were assessed for two baseline and four intervention sessions. We applied a mixed-method design, combining advanced sensors methodology, observations, peer nominations, self-reports, qualitative field notes and an interview with the playground teachers. Results: Findings indicated for all children a decrease during the intervention in social interactions and social play and no change in network centrality. Children without disabilities displayed also an increase in solitude play and in the diversity of interacting partners. Enjoyment of LPP was high for all children, yet children with disabilities did not benefit socially from the intervention and became even more isolated compared with baseline level. Conclusions: Social participation in the schoolyard of children with and without disabilities did not improve during LPP in a mainstream setting. Findings emphasize the need to consider the social needs of children with disabilities when designing playground interventions and to re-think about LPP philosophy and practices to adapt them to inclusive settings and goals.

Social connectedness at school is crucial to children's development, yet very little is known about the way it has been affected by school closures during COVID-19 pandemic. We compared pre-post lockdown levels of social connectedness at a school playground in forty-three primary school-aged children, using wearable sensors, observations, peer nominations and self-reports. Upon school reopening, findings from sensors and peer nominations indicated increases in children's interaction time, network diversity and network centrality. Group observations indicated a decrease in no-play social interactions and an increase in children's involvement in social play. Explorative analyses did not reveal relations between changes in peer connectedness and pre-lockdown levels of peer connectedness or social contact during the lockdown period. Findings pointed at the role of recess in contributing to children's social well-being and the importance of attending to their social needs upon reopening.

Detecting and analyzing group behavior from spatio-temporal trajectories is an interesting topic in various domains, such as autonomous driving, urban computing, and social sciences. This paper revisits the group detection problem from spatio-temporal trajectories and proposes “WavenetNRI”, a graph neural network (GNN) based method. The proposed WavenetNRI extends the previously proposed neural relational inference (NRI) method (an unsupervised learning approach for inferring interactions from observational data) in two directions: (1) symmetric edge features and edge updating processes are applied to generate symmetric edge representations corresponding to the symmetric binary group relationships; (2) a gated dilated residual causal convolutional (GD-RCC) block is adopted to capture both short and long dependency of the edge feature sequences. We evaluated the performance of the proposed model on three simulation datasets and three real-world pedestrian datasets, using the Group Mitre metric to measure the quality of the predicted groups. We compared WavenetNRI with four baseline methods, including two clustering-based and two classification-based methods. In these experiments, NRI and WavenetNRI outperformed all other baselines on the group-interaction simulation datasets, while NRI performed slightly better than WavenetNRI. On the pedestrian datasets, the WavenetNRI outperformed other classification-based baselines. However, it did not compete against the clustering-based methods. Our ablation study showed that while both proposed changes cannot be effective at the same time, either of them can improve the performance of the original NRI on one dataset type.