Multi task processing AI system design for autonomous vehicle driving | |
| Author | Bondalapati, Sahrudai |
| Call Number | AIT Thesis no.ISE-19-40 |
| Subject(s) | Machine learning--Design Autonomous vehicles--Technological innovations Artificial intelligence--Engineering applications |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Microelectronics and Embedded Systems, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | |
| Abstract | Nowadays, machine learning has found a wide range of applications in many fields. Deep learning techniques especially Convolutional Neural Networks and Recurrent Neural Networks are most commonly applied to analyzing visual imagery. Face recognition, image recognition, object recognition etc., are the major applications of the CNN. Autonomous vehicles are one of the most important use of machine learning as it has the potential to save millions of lives. The proposed model describes the development of a multi-processing AI system design which can detect and classify humans, vehicles and other objects, detect lanes and also predict steering angle and speed from a road image. A combined model based on 3D network and lane detector, YOLOv3 and lane detector are the architectures used in this work. We are using a part of udacity dataset, which contains image frames from driving videos with information of steering angle and speed for each frame. The YOLOv3 network is pre-trained on coco dataset and the lane detector is pre-trained on lane detector dataset. The proposed models detect humans, vehicles and other objects using bounding boxes and further classifies them respectively. The combined model based on 3D network and lane detector predicts the steering angle and speed. The lane detector detects the lane. The experimental results indicate that the proposed CNN architectures are efficient and very successful network configurations for predicting steering angle and speed. |
| Year | 2019 |
| Corresponding Series Added Entry | |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Industrial Systems Engineering (ISE) |
| Chairperson(s) | Mongkol Ekpanyapong,;Keun, Song Weon (Co-chairperson) |
| Examination Committee(s) | Keun, Song Weon;Bohez, Erik L.J.;Huynh, Trung Luong; |
| Scholarship Donor(s) | Asian Institute of Technology Fellowship; |
| Degree | Thesis (M. Eng.) -- Asian Institute of Technology, 2019 |