A scene recognition, pedestrian detection and vehicle action detection model, with an extensive dataset encapsulation API. The API is written for proprietary data by Daimler.
A web application in Django that takes a video and corresponding CAN data from Daimler's dataset, and processes them, shows the object detection processed video with synced CAN as a slider
NOTE: This code is useless for anyone except the sponsors of the project who have the coresponding data.
$ sudo docker run -it --rm\
--name scener2\
-p 8000:8000\
dumbpy/scener2:firefox
access from firefox at localhost:8000. The sliders don't work on chrome and will be fixed in future updates.
Under the Supervison of Prof. Manjesh Hanawal1, Prof. P Balamurugan1 and Dr. Tilak Singh2
1Indian Institute of Technology Bombay 2Mitsubishi Fuso Truck and Bus Corp. Japan