Computer Vision Intern (Summer)
What you will do
Imec-USA is looking for a graduate student intern in the area of computer vision who want to apply machine learning/AI on new imagers and sensors. Millimeter-wave (mmW) radar is a very useful sensor for autonomous navigation due to its low cost, long-range detection, and undisrupted performance by harsh weather conditions. Typically, radar resolution is low, therefore it is used in conjunction with other sensors such as RGB cameras and Lidar that delivers higher resolutions in order to discriminate objects of interest that can be used for navigation and/or taking decisions. In this project, you will be exploring the use of radar as a solo sensor (instead of multiple sensors) in the tasks of object recognition and 3D semantic segmentation. To achieve the goal, you will be by levering on I/A to learn to obtain dense radar reconstructions of common objects on streets.
Your responsibilities include:
- Generate a dataset using our internal simulation tool to create virtual scenes from CAD models. These CAD virtual scenes will be interpreted as groundtruth.
- Create one or several deep network that learn to use the non-ideal reconstructions from radar and antenna signals as input data, to create output that is as close as possible to the original 3d CAD models.
- Explore several data representations approaches that can deal with 3D data such as:
- Higher resolution images from multiple projections (top view, frontal view, lateral view)
- Voxelize + 3D convolutions
- Use the voxels in a similar way to PointNet 
- You must select the most appropriated ones for the problem and implement them.
- Train and test several deep networks and compare them.
- Lead or help lead a corresponding publication about the work.
 PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation. R. Qi Charles ; Hao Su ; Mo Kaichun ; Leonidas J. Guibas. CVPR 2017
What we do for you
In exchange for your talent, passion and expertise, you will join an ambitious organization, with challenges there for the taking. This internship offers you the opportunity to work for imec, a globally recognized leader in high frequency electronics design. With the guidance of senior researchers, you will participate in the creation of an algorithm that potentially could be available in one of our sensors. You will have an assigned mentor to guide you through this highly valuable, hands on learning experience. You will work with top researchers from imec and from our industry partners. We are a not for profit research institution where you can work across a variety of projects and topics. We work close to industry, so your expertise and efforts often convert to solutions that are widely utilized and make positive impacts in the world. Our progressive and informal working environment offers you a range of possibilities to take initiative. This is your opportunity to contribute to the technology that will determine the society of tomorrow. Your valuable contribution and that of your colleagues make imec a top player in its field.
Who you are
- Currently enrolled in or a recent graduate from a relevant Master’s or PhD program in a top university in a relevant discipline such as Computer Science, Computer Engineering, or Electrical Engineering.
- Experience tinkering with deep learning frameworks.
- Experience with software programming.
- Experience with signals, conditioning, acquisition, and processing.
- Experience applying mathematical concepts to physical phenomena.
- Highly motivated to work in a research environment.
- Excellent communication skills and ability to work in a multidisciplinary team.
It is the policy of imec to provide equal employment opportunities to all employees and employment applicants without regard to unlawful considerations of race, religion, color, national origin, ancestry, sex, sexual orientation, gender, gender identity or expression, genetic information, age, disability, medical condition, marital status, military or veteran status or any other classification protected by applicable local, state or federal laws. This policy applies to all aspects of employment, including, but not limited to, hiring, job assignment, compensation, promotion, benefits, training, discipline and termination. Reasonable accommodations of disability and religion are available for qualified individuals, upon request, provided that such accommodation does not present an undue hardship to imec.