- Responsible for the research and development of image processing and pattern recognition algorithms, such as image enhancement, object detection, defect detection, object classification, etc;
- Responsible for the research and development of machine learning technology, such as feature extraction, (un)supervised learning, deep learning, etc.;
- Responsible for the software development & programming with the algorithms, including algorithm optimization, coding, verification, product development and testing in both Windows and Linux;
- Responsible for user interface (UI/UX) design;
- Resolve technical problems relative to software and system platform.
- Master's degree holder with 3+ years relevant experience or PhD holder in Computer Science, Maths and Physic and EE or related disciplines. Those with less experience will be considered as Engineer;
- Experience in computer vision, image processing or pattern recognition is a must;
- Experience in machine learning/deep learning algorithm is a plus;
- Familiar with computer programming skills on C/C++, OpenCV, Matlab/Octave, OpenGL, etc is preferred;
- Experience in programming of Linux/QT/.Net/Python/Perl, Linux OS maintenance, SVN/Git, OpenCL/CUDA is a plus;
- Good command of written and spoken English, self-motivated and good communication skills.
Appointment will be on renewable contract terms with a competitive salary and performance-linked variable pay. Fringe benefits include paid leave, medical and dental benefits, insurance coverage and contribution to MPF. The incumbent will normally work under a five-day week schedule. Interested candidates should send application (quoting Ref. No.) with detailed resume and, current and expected salary to the HR Department by email (preferable) or post no later than 15 October 2016. Email: [email protected] Post: 5/F, Photonics Centre, 2 Science Park East Avenue, Hong Kong Science Park, Shatin, Hong Kong. Only short-listed candidates will be notified. Personal data provided by applicants will be used for recruitment purposes only.