News
Media coverage, awards, papers, and milestones from the MOTIF Lab.
- Accepted paper
Alina Hagen's street-view accessibility paper accepted to WACV 2027
"Do VLMs See What Sensors Feel?" — an expert-guided design that scores wheelchair accessibility from Google Street View and checks it against GPS-derived wheelchair dwell behavior — will be presented at IEEE/CVF WACV 2027.
- Accepted paper
DriveMotion accepted to IEEE/CVF WACV 2027
Our large-scale, multi-source benchmark for driver motion sequence modeling and forecasting will be presented at WACV 2027 in Lake Buena Vista, Florida (January 2027).
- Award
CoRL 2026 Travel Award for "Observe Before You Alert"
Yuhang Wang received a CoRL 2026 Travel Award for the VLAlert paper on adaptive driver alerting with vision–language models.
- Accepted paper
Seven lab papers accepted for presentation at TRB 2027
Five first-authored by Yuhang Wang — TriDrive (CockpitWM), Observe Before You Alert, BATON, Human Takeovers in Level-2 ADAS, and DriveDNA — plus two co-authored papers will be presented at the Transportation Research Board Annual Meeting in January 2027.
- Preprint
TriDrive: a real-time cockpit world model, now on arXiv
TriDrive jointly forecasts driver kinematics, vehicle dynamics and road demands from one automation-conditioned transition model, runs at 5 Hz on a comma four, and was validated in a 14-participant on-road study.
- Accepted paper
"Observe Before You Alert" accepted to CoRL 2026
VLAlert casts driver alerting as a tri-action policy — SILENT, OBSERVE, ALERT — and releases VLAlert-Bench: 13,534 videos and 192,892 labelled one-second ticks from six driving-event datasets.
- Preprint
DriveMotion is on arXiv and Hugging Face
A multi-source, skeleton-only benchmark for driver motion forecasting: 400 hours, 360 drivers, 133 keypoints at 10 Hz, released under CC BY 4.0 (gated).
- Media
USF News, FOX 13, and ABC7 Suncoast cover the lab's AI dashcam for post-hurricane road assessment
Three Florida outlets reported on the lab's AI dashcam prototype, which automatically flags potholes, flooding, downed trees, and damaged signs so agencies can assess storm-damaged roads in hours instead of days — most recently ABC7 Suncoast, on the Sarasota County deployment.
- Accepted paper
Three papers accepted to IEEE ITSC 2026
ADAS-TO, cut-in gap acceptance toward autonomous vs. human-driven vehicles, and a closed-loop multi-agent framework for passenger load estimation will be presented at the IEEE International Conference on Intelligent Transportation Systems 2026.
- Preprint
New preprint: Scale-CDA brings AI-assisted cooperative driving to production cars for under $1,000
A new lab preprint introduces Scale-CDA, an open-hardware/open-software toolchain that retrofits ordinary production cars for generative-AI-assisted cooperative driving automation using off-the-shelf parts costing under US $1,000.
- Preprint
DriveDNA: driving-style identification without the shortcuts
4,121 drives from 465 pseudonymous drivers across 115 vehicle models (975 hours), with a benchmark built to separate driver behavior from vehicle, route and condition shortcuts.
- Milestone
MOTIF Lab website launched
motiflab.net is live: research threads, team, publications and open datasets from the MOTIF Lab at the University of South Florida.
- Preprint
BATON: benchmarking bidirectional human–automation control transitions
A multimodal benchmark of handovers and takeovers in naturalistic driving-automation routes; the July 2026 v2 reports 781 routes, 173 drivers and 204.9 hours, with a frozen benchmark of 565 routes and 3,593 events.
- Accepted paper
Three papers accepted for presentation at TRB 2026
Including an empirical characterization of electric-vehicle acceleration and its implications for lane changes and bottleneck capacity, presented in Washington, D.C., in January 2026.
- Accepted paper
"From OpenLKA to LKAlert" accepted at INFORMS 2025
An abstract on moving from the OpenLKA dataset to predictive lane-keeping alerts was accepted to the INFORMS Annual Meeting 2025 (Intelligent Transportation Systems and Cybersecurity session) in Atlanta.
- Accepted paper
OpenLKA accepted to IEEE ITSC 2025
The open dataset of lane keeping assist from production vehicles under real-world driving conditions was presented at the IEEE International Conference on Intelligent Transportation Systems 2025 in Gold Coast, Australia.
- Accepted paper
Extended abstract accepted for a poster at the Second USF AI + X Symposium
An extended abstract from the lab on driving-automation data was accepted for a poster at the Second USF Artificial Intelligence + X Symposium (October 2024).