Research Areas

Our interdisciplinary research teams work at the intersection of machine learning, computer vision, natural language processing, and AI safety to develop systems that advance autonomous driving, healthcare AI, and AI dependability.

Computer Vision & Autonomous Driving
Advancing vision-language fusion for real-time autonomous driving with state-of-the-art performance in navigation and perception.

Current Projects

XYZ-Drive: Vision-Language Fusion for Autonomous DrivingCompleted

A single vision-language model achieving 95% success rate on autonomous driving benchmarks, surpassing previous state-of-the-art by 15%.

View paper
NovaDrive: Early Goal-Guided Multi-Scale FusionCompleted

A single-branch architecture that raises success rate to 84% and reduces collision frequency from 2.6% to 1.2%.

View paper
PhysNav-DG: Robust VLM-Sensor FusionCompleted

A novel framework integrating classical sensor fusion with vision-language models for transparent navigation.

View paper
Healthcare AI & Conversational Agents
Developing embodied conversational agents and multimodal systems for healthcare applications and therapy.

Current Projects

Dementia Reminiscence Therapy with ECACompleted

An embodied conversational agent system delivering personalized reminiscence therapy for people with dementia.

View paper
GenECA: Real-Time Multimodal Conversational AgentsCompleted

A general-purpose framework for real-time adaptive multimodal embodied conversational agents.

View paper
Tri-Modal Depression ClassificationCompleted

Integrating large language models into multimodal architecture for automated depression classification.

View paper
AI Safety & Dependability
Ensuring the safety and reliability of AI systems in cyber-physical environments through verification and validation frameworks.

Current Projects

DURA-CPS: Multi-Role Orchestration for AI AssuranceCompleted

A novel framework employing multi-role orchestration to automate the iterative assurance process for AI-powered cyber-physical systems.

View paper
WebNav: Voice-Controlled Web NavigationCompleted

An intelligent agent for voice-controlled web navigation, improving accessibility for visually impaired users.

View paper
Applied ML & Accessibility
Translating cutting-edge ML research into practical applications that improve accessibility and user experience.

Current Projects

CLIP-MG: Micro-Gesture RecognitionCompleted

A pose-guided semantics-aware CLIP-based architecture for micro-gesture recognition on the iMiGUE dataset.

View paper
Semantic Web for UI Element AnnotationCompleted

Leveraging semantic web technologies to improve automated annotation of UI elements for better accessibility.

View paper

Research Highlights

Explore some of our most impactful recent research projects and publications.

Autonomous Driving
XYZ-Drive: 95% Success Rate

Vision-language fusion for real-time autonomous driving, achieving 95% success rate and surpassing previous state-of-the-art by 15%.

Read the paper
Healthcare AI
Dementia Therapy with ECA

A personalized embodied conversational agent system delivering reminiscence therapy for people with dementia.

Read the paper
AI Safety
DURA-CPS: AI Assurance

Multi-role orchestration system for ensuring the safety and reliability of AI in cyber-physical systems.

Read the paper