What is this?
Deep Rectified System for High-speed Tracking in Images is an open-source object detection API that can be used by anyone free of cost to perform object detection on normal images and vehicle detection in aerial images. Its aimed at the generic user who wants to step into the A.I based computer vision domain and the portal helps to form a basic intuition of the possibilities this field offers.
Featured Demos
Object Detection for normal scenes
- Detect objects in everyday life
This model uses a Faster RCNN trained on 37 classes using the PASCAL VOC 2012 dataset. It can detect objects like cars, buses, people, traffic signs, plants etc. The user need to upload an image on the portal in order to get bounding boxes around the objects which the model has identified. To take a look at themodel, click the link below.
Check out Demo
Project Title
- Vehicle Detection in Aerial Images
Vehicle detector model for aerial images is an F-RCNN trained on VEDAI aerial image dataset. The model can detect and and classify vehicles from aerial images in very challenging conditions and was used to prove the Faster-RCNN's small target detection capability.
Check out Demo