Maneuvering Control based on Intention Prediction through Visual Perception of Traffic Network Participants
Eng. Mohamed Manzour
The video shows the pedestrian's detected motion state (Standing/Walking). This detection is obtained using a Convolutional Neural Networks (CNN) model which is trained on the Joint Attention in Autonomous Driving (JAAD) dataset ([ Ссылка ]).
This model takes the cropped bounding box of the pedestrian as input in order to obtain the pedestrian's motion state.
The frame rate of this video is 30 frames per second.

For more information, kindly check Section x in Chapter y in the following Master of Science in Mechatronics Engineering Thesis:

Please, visit the research group website for more information about the academic and research activities [ Ссылка ]

свернуть