Request Invitation

RoadSense is the only widely available cloud-based solution for full road scene understanding. It understands complex road scenes and generates rich metadata describing their content. With custom state-of-the-art computer vision and deep learning models trained on thousands of driving miles, RoadSense delivers best-in-class analytics directly from uncalibrated video.

RoadSense contains a suite of analytics that detect and recognize semantic roadway infrastructure details in both urban and highway settings. RoadSense understands traffic entities such as road signs and signals, intersections, lanes, and painted markings. It can also detect vehicles, perform fine-grained recognition such as make, model, and color classification, and recognize vehicle density and behaviors.

RoadSense is continually evolving. We will continue to deploy new RoadSense analytics to our cloud platform, and we regularly upgrade our existing analytics with new models trained on our growing database of road scene video.

CitySense is a suite of state-of-the-art analytics for city-level forensics.

CitySense can ingest and index large collections of surveillance video and efficiently perform large-scale targeted search queries to retrieve, for example, vehicles or persons of interest. It also understands criminal activity and can automatically detect and recognize many classes of criminal actions across video modalities such as surveillance, dashcam, and body-worn cameras.

CitySense is continually evolving. We will continue to deploy new CitySense analytics to our cloud platform, and we regularly upgrade our existing analytics with new models trained on our growing database of city-level video.

Our experienced team of computer vision engineers and scientists is available to develop and deliver custom software solutions for your video understanding needs.

Our custom solutions can be integrated into your existing workflows, or we can deliver them to you as a convenient end-to-end system through an on-premises deployment of our computer vision platform. Our vision engine is highly extensible and generalizable to new analytics, so we can fulfill your computer vision needs efficiently and elegantly.

Example projects include automated or semi-automated data annotation services, training and deploying custom machine learning models on your proprietary data, and comparative evaluation of your existing data analytics.

Contact us at solutions@voxel51.com to set up a call.