Visualization

Platform design for teleoperated social robots in classrooms, enabling safe child-robot interactions and iterative prototyping of educational robot applications.

SCOPE integrates natural language processing with camera control for edge deployment, executing perception and planning locally without cloud dependencies.

Pipeline using multiple language models and consensus voting to reduce manual effort in systematic literature review screening, with human oversight through visual analytics.

Automated pipeline for converting Excel spreadsheets to dynamic, publication-quality visualizations using machine learning chart recommendation and interactive dashboard functionality.

Deep learning framework quantifies craniofacial growth patterns and sexual dimorphism from cephalometric radiographs using automated feature extraction and saliency mapping without manual annotation.

Analysis of image types and multimodal slide-speech relations in Three Minute Thesis presentations by engineering PhD students addressing non-specialist audiences.

Narrative Scaffolding positions narrative construction as the primary interface for data-driven exploration, enabling deeper reflection and broader investigation patterns while preserving.

Deep learning model using ResNet-50 classifies fibro-osseous jaw lesions from histology images with 86% accuracy, outperforming pathologists in differentiating fibrous dysplasia, cemento-ossifying.

Graph matching framework for cross-modality intravascular ultrasound and optical coherence tomography sequence registration with simultaneous temporal and rotational alignment.

System mapping methods were used to visualize PCORnet's data architecture, revealing common elements and heterogeneity across CRNs to support study design, stakeholder dialogue, and learning.