System improves safety and transit reliability by enabling automated enforcement of bus lanes, bike lanes, and bus stops.
Dangerous traffic violations like double parking or illegal parking in bike lanes, bus lanes, or at bus stops are difficult to process at scale with traditional photo enforcement, which relies on manual review of multiple photographs in order to find a potential traffic violation and transcribe the license plate of the offending vehicle.
Hayden AI’s mobile perception platform utilizes computer vision and deep learning technology that automatically detects context features to determine if a traffic violation occurred in video evidence. By inputting video frames of a potential traffic violation into a deep learning algorithm, Hayden AI’s solution “scores” each evidence package to identify difficult cases. Retraining the AI algorithms using these difficult instances leads to significant improvement in performance and establishes a positive feedback loop of continuous improvement of the AI system.
Also Read: SiLC Secures $25 Million in Additional Funding to Advance Machine Vision for AI Applications
A strong team of Hayden AI staff invented this innovative system, including Wiktor Muron, Maciej Budys, Andrei Liaukovich, Marcin Grzesiak, Michael Gleeson-May, Shaocheng Wang, Morgan Kohler and Chief Technology Officer Vaibhav Ghadiok.
“Using advances in AI to determine if traffic violations occurred gives our public sector partners the ability to solve large, long-standing problems with their existing resources,” said Vaibhav Ghadiok, CTO of Hayden AI. “Keeping bus lanes, bike lanes, and bus stops clear for their intended users is essential to improving transit service reliability and safety for everyone on the road.”
Hayden AI is the US market leader in mobile automated bus lane enforcement and bus stop enforcement, with hundreds of camera systems installed on transit buses across the country.
SOURCE: BusinessWire
