Home | Stanford Translational AI (STAI)

Stanford Translational AI (STAI) Lab

The Stanford Translational AI (STAI) Lab unites interdisciplinary researchers from fields such as Artificial Intelligence (AI), Computer Vision (CV), Neuroscience, Medical and Neuro Imaging, and Psychiatry, with the goal of addressing critical challenges in Precision Healthcare and Mental Health. Our mission is twofold: to drive innovation both in AI methodologies and in their real-world clinical applications. Our group consistently publishes cutting-edge research in top-tier AI and CV conferences, while also translating these technologies into impactful clinical applications that are featured in leading medical and interdisciplinary journals.

Follow the lab

Projects

Computer Vision
3D Digital Humans
Our project on human rendering from occluded videos addresses the challenge of accurately reconstructing 3D human bodies from single-view, in-the-wild video data, where unexpected occlusions often occur.
Read more →
Ambient Intelligence (AmI)
Detecting Neuropsychiatry Symptoms (NPS)
Detecting neuropsychiatric symptoms (NPS) such as anxiety, agitation, or depression is essential because it enables the early identification and more tailored treatment of mental health conditions. Currently, NPS is detected using patient- or caregiver-reported outcomes, which can be inherently subjective and prone to various biases.
Read more →
Ambient Intelligence (AmI)
AmI in Hospitals (ICU)
Our research project focuses on leveraging ambient intelligence and computer vision within intensive care unit (ICU) environments to support clinical documentation and patient monitoring.
Read more →
Ambient Intelligence (AmI)
Senior Care
Our senior care project focuses on developingadvanced AI models to improve the quality of life and safety for older adults. By analyzing data from ambient sensors, our AI models can track seniors’ daily behaviors and detect any unusual changes or immediate risks. This system provides family caregivers with timely alerts about potential issues, offering peace of mind while supporting seniors’ independence. Through this innovati…
Read more →
Computer Vision
Video Understanding & Motion Analysis
We are working on a series of works around Video Understanding & Motion Analysis, focusing on developing models to understand human motion, activities, and behaviors in real-world environments from videos. This work includes building systems for video action recognition and analysis,
Read more →
Computer Vision
Activity & Gesture Analysis
We explore video language models for understanding human activities, focusing on detailed and fine-grained actions, including the activities of daily living (ADLs) in home environments. We developed InteractADL, a multi-view video dataset with scene graph annotations and atomic action labels to improve computer vision models' ability to understand ADLs.
Read more →
Computational Neuroscience
Brain World Model
Despite the remarkable advancements in neuroscience and related fields, we still cannot fully explain the complexities of the human brain. Foundation models hold promise, but so far, these applications have focused only on single modalities of data, such as MRI, or one of its functions, such as vision.
Read more →
Computational Neuroscience
Neural Decoding
Neural decoding is the process of understanding how brain activity corresponds to different stimuli. It is often considered one of the primary objectives in cognitive sciences.
Read more →
Mind & Motion
Joint Video-MRI analysis
We aim to employ recent advances in computer vision and artificial intelligence to assist in differentiating dementia phenotypes with a solid understanding of the neural underpinnings of gait and mobility disruptions related to dementia. Our analysis will identify personalized ...
Read more →
Mind & Motion
Neural mechanisms of gait impairment in PD, AD/ADRD, & MCI
Motor symptoms, such as changes in gait and functional mobility, are among the earliest detectable signs of various neurodegenerative diseases, including Parkinson's and Alzheimer's. However, our understanding of the underlying causes—specifically, which brain changes are associated with specific phenotypes—is limited.
Read more →
AI Trustworthiness
Confounders in Medical Studies
The presence of confounding effects (or biases) is one of the most critical challenges in using deep learning for medical imaging studies. Confounders influence the relationship between input data (e.g., brain MRIs) and output variables (e.g., diagnosis), often leading to spurious and biased associations.
Read more →
Mental Health
Estimating levels of depression and anxiety using multimodal AI
Nearly 60 million people in the United States struggle with mental illness each year. Yet, healthcare providers identify depression in only 47% and detect clinical deterioration in only 21% of cases.
Read more →
Mental Health
Foundation Models & LLMs in Psychiatry
This project focuses on developing and creating rigorous evaluation frameworks for multimodal foundation and large language models to advance mental health research and clinical applications. Leveraging advanced AI techniques, we integrate diverse data types—including text, imagery, and behavioral data—to develop tools that support psychiatric diagnosis, prognosis, and personalized treatment. Our innovative approach aims to br…
Read more →
Mental Health
AI-assisted personalization of cognitive training sessions for order adults
Computerized cognitive training (CCT) is a scalable, well-tolerated intervention that has promise for slowing cognitive decline. The effectiveness of CCT is often affected by a lack of effective engagement. Factors like Mental fatigue and reaction time are the primary factors for compromising effective engagement in CCT, particularly in older adults at risk for dementia.
Read more →