For immediate placement, we are currently looking for an Engineer (m/f/x) specialized in image processing.
The main task will consist in developing and implementing cutting edge algorithms in order to improve the image database. Your goal will be to create mission-specific image datasets for training deep convolutional neural networks using image processing and SQL database management tools. This task includes the creation of original tools for image acquisition, automatic segmentation, annotation and data augmentation technics.
Your new tasks and responsibilities:
- You will be responsible for organising, developing and testing the image database:
o Action focused on the implementation of a new architecture able to manage the object sub-parts
- Improvement of the existing image preprocessing tools for automatic object segmentation, bounding box automatic detection and automatic contrast and luminosity correction
- Implementation of data augmentation technics (homothety, rotation, translation, mosaic…)
- To export training, validation and test datasets in order to benchmark the trained deep networks.
- Training and evaluation of deep neural networks
- Writing and distribution of the work advancement status
- Supervision of technicians and interns
- You are an engineering graduate or you have a Master's degree (CTI certified)
- Knowledge in the following areas is required:
o Skills in datascience, computer vision and AI
o Programming languages : SQL, Python, Qt, C++
o Deep learning and databases
- Good communication skills. You like to work in a team and to take initiatives. You are an effective organiser, easily adapting to new situations.
- Very good command of English
What we offer: Fixed-term contract for one year.
ISL offers attractive remuneration, flexible work arrangements and many opportunities for vocational training and further development. The position is open to people with disabilities.
If this challenging description fits your profile, we are looking forward to receiving your complete application.
Published on 24 May 2022