Object Recognition, Localization and Tracking Engineer
DOT Technology Corp.
Regina, SK, Canada
Object recognition is an essential technology used in autonomous vehicles where the main goal is understanding scene context and detecting instances of objects in image frames. Different approaches exist for object recognition including computer vision (for feature extraction), machine learning (for classification) and deep learning (for black box approach) techniques. After recognition and detection, localization and tracking in the image frame is required which can be achieved using common computer vision techniques. Ultimately, the successful applicant has a strong background in computer vision, mathematical optimization and deep learning as well as a basic understanding of the common sensors used on autonomous vehicles. The main goal for this position is to help develop a software module to recognize scene context, detect objects in the image frame and localize and track the objects using the information available from a multitude of sensors including LiDAR, camera, radar, ultrasonic, GPS and etc. This software module will help the machine to operate in the field safely.
We would like you to have:
- Bachelor’s degree in engineering or computer science with hands-on experience. Master’s degree or PhD preferred.
- Strong background in optimization, linear algebra, computer vision and machine learning
- Strong programming skills in Python or C/C++
- Hands-on experience working with computer visions techniques and tools
- Hands-on experience working with deep learning frameworks (TensorFlow, Caffe, and etc.)
- Experience with common sensors on autonomous vehicles (LiDAR, radar, ultrasonic, and etc.)
- Experience developing embedded firmware for safety-critical systems in production environments
- 2+ years of relevant work experience
It would be great if you also had:
- Familiarity with CAN bus and GPS
- Hands-on experience building robots!
In this role you will:
- Assess project requirements, undertake relevant research, and implement new and/or utilize existing algorithms
- Create testing procedures, simulation models, and write test driven code
- Manage, train, and tune state-of-the-art deep neural networks
- Source, evaluate and test sensors
- Test the product in controlled and real world situations
- Collect, process, and manage training data, including training and testing pipelines
- Continual improvement and management of DOT autonomous driving systems
- Ensure safe operation of DOT