DEEP LEARNING SCIENTIST HF

Job description

Context :
Part of EDF Energy’s Research & Development centre, the Digital Innovation team is young, dynamic and ultra-curious. We are currently 12 people: developers, data scientists, designers and innovation experts. Our job is to not only keep a beady eye on the most cutting-edge digital technologies, but also to turn those technologies into proof of concepts that show how they could bring value to the business. As such, we operate like an in-house startup, but one that works on virtual and augmented reality, mobile apps, data visualisation, artificial intelligence, blockchain, bots, connected home and IoT... all wrapped in a beautiful layer of great user experience and visual design.
You’ll be equipped with a MacBook Pro, Slack, Google Apps, access to your own favourite tools (you choose), a collaborative open environment and an endless supply of new ideas and technologies. We are constantly learning: monthly half-day skill sharing sessions are supplemented with hack days, talks from academic researchers and meetups. There might be opportunities to visit EDF innovation centres for various events in the UK and abroad (recent ones include hackathons in Paris and Beijing).
Function :
The objective of the VIE, as a deep learning scientist, is to contribute to existing and new data science and digital projects, in collaboration with Customer and Power Generation business units. The VIE will be collaborating with developers and data scientists of the R&D Digital Innovation team as well as EDF Group R&D based in Paris, France. This would include, for example, contributing to on-going projects such as object detection in panoramic images of a power plant for its maintenance or decommissioning or defect-detection on wind turbine blades. This will also include exploring new projects such as, for example, machine learning for physical simulations or advanced text mining of our customer interactions.
You will work on the following topics:
- Developing deep learning algorithms, primarily for advanced computer vision
- Experimenting and implementing state of the art convolutional deep neural networks
- Experimenting state of the art deep learning processings: time series analysis, advanced text mining or physical modelling
- Data formatting and conversion
- Development on GPU or distributed environments
- Collaborating with developers and data scientist to deliver integrated projects
- Innovation and technology monitoring, looking for applications of recent research
Skills
The person appointed should be able to demonstrate strong analytical skills along with a great level of English.
Required skills and experience:
- Applied deep learning experience in computer vision (object recognition, image segmentation, instance segmentation)
- Strong math background, with extensive understanding of machine learning, geometry and probabilities
- Strong experience in Python, C++ is a plus
- Strong experience in CUDA, TensorFlow
- Keras
- Ability to write clear and well maintained code
Desirable skills and experience
- Experience in building or evaluating a large dataset
- Experience in laser scanning and 3D modelling are a plus
- Experience in Unity3D and C# are a plus
- Author of a publication in a journal or conference in the related field
We give the greatest importance to being innovative, flexible and being keen to work within a team animated by a
start-up spirit. An interest in energy and the energy industry would be an asset. The VIE mission will be taking place between Brighton and London.
Examples of our work
About EDF Energy R&D: https://www.edfenergy.com/about/research-development
An example of our work: https://www.youtube.com/watch?v=J885N-dXZfU
One of our hackathons: https://youtu.be/JT1rsJIUWkc