gnanesh kundarnahalli mahadevanna

Topic: "Generation of Synthetic Fisheye images of Indoor environments in the context of activities in daily living". • Generation of synthetic datasets using Unity Game Engine.• Conversion of the Theodore dataset to AVA dataset format using Python.• Implementation of the Slow-Fast (CNN-based) model using RESNET.• Training and Validation of the model for the Spatio-temporal action recognition.

• Data labeling of LIDAR images, projection to RBG images and interpolation of data setused for the detection of people outside and inside the car is used.• Outside, detection takes place in the vehicle's surroundings, where people and otherobjects are detected and evaluated for driver assistance and automated driving.• Testing and Evaluation: Bug detection in the data labeling tool