Heavy vehicle detection

Our model has been custom trained using thousands of truck images and is built on the powerful YOLO model that allows for real-time object detection.

• Detect heavy-vehicles in an instant.
• Get a photo to corroborate the report.
• Read vehicles' licence plates and a timestamp.
• Prevent heavy-vehicle accidents taking place in urban roads.

Day sample

Our model, as you can see here work perfectly on Indian roads full of traffic, in fact, it is optimised for them!

• Detects vehicles on all sorts of roads.
• Works very well on roads full of traffic.

Night sample

Our model, as you can see below, has also been primed to detect heavy-vehicles in shabby light conditions like at night-time.

• Works at night.
• Works under any lighting conditions.

Road Detect

We worked on a consumer-facing model ported on mobile devices to help reduce cycling accidents. Our iOS application is designed to detect potholes in real-time when mounted on a cycle with the camera facing the road.

• Detect potholes
• Be warned of approaching dangers on the road.
• Take approapriate action.
• Prevent accidents.

Technical Details

We use SOTA algorithms of object detection and tracking to complete our software. The SSD MobileNet model is trained for over 30,000 images of 4 types of potholes. We integrated this SSD model into an app and then are in the process of using this app to register potholes for local municipal coporations. The model works with a mAP of 60% and is definitely going to benefit both the government with automatic road analysis and the general public with avoiding accidents. We used the the TF converter to create an edge device compatible model and adapt realtime detections on a phone. The project can help governments significantly reduce accidents and better register pothole data.

Beta model in action

To the right, you can observe the beta version of our quantized SSD MobileNet model ported on iOS in action.

Once a bounding box is created, an alert is sounded. This feature is aimed towards cyclists who may otherwise not pay attention to a pothole detected.

We will be releasing Road Detect on the App Store soon.