Using the Indian Driving Data set we have trained a Faster RCNN model using the Transfer Learning and got a good MAP for the same.
->Unstructured environment is an environment that contains many obstacles and where vehicle localization is difficult. ->Most natural environments in India are unstructured. ->Scenarios like dense traffic condition and places where there are many obstacles ,detection in such environment is difficult. ->There is no lane information to guide or constrain the actions of the vehicle,current algorithms don’t generalize well when tested on diverse data distributions. ->It is a challenging task in unstructured environments where lanes vary significantly in appearance and are not indicated by proper markers.
Indian Driving Dataset (IDD) - In recent years several datasets are available for autonomous navigation tend to focus on structured driving environments which corresponds to well-delineated infrastructure such as lanes, a small number of well-defined categories for traffic participants, low variation in object or background appearance and strong adherence to traffic rules. IDD, is a novel dataset for road scene understanding in unstructured environments. It consists of 10,004 images, finely annotated with 34 classes collected from 182 drive sequences on Indian roads.
- Replace the drive path with user drive path.
- Set the number of epoch you want to train the model.
- Run the last cell to get all performace parameters.