Cédric Rommel, PhD

Machine Learning Researcher

Open source

I believe collective progress is accelerated by open-source practices in code, science and knowledge in general. Hence, I enjoy collaborating to open-source libraries such as the ones below. You can also check my github for other open-source codes.

Braindecode

Braindecode bridges the worlds of deep learning and neuroscience based on EEG signals. Besides cleaning a few things here thhere, my core contribution up to now has been to include the data augmentation module for EEG data.

Skorch

Skorch is a user-friendly wrapper of the neural newtork python library Pytorch, which follows the worldwide recognized API principles of scikit-learn (developped by the Parietal team at Inria).

Proprietary software

I also had the chance to contribute to beautiful proprietary software during my professional trajectory:

Ava

While speech technology has incredibly improved in the new era of deep learning, speech technology has mostly been used for cool but gadget applications. Ava is the best exception to this rule that I know, as it can help 450M deaf and hard-of-hearing people worldwide live a fully accessible life. It is based on state-of-the-art speech recognition and a unique real-time speaker identification technology that I have had the opportunity to help building. Try it out, it’s free!

OptiClimb

Air traffic accounted for 3% of global emissions in 2017 and the climb phase is the most emitting and fuel consuming part of a flight. Moreover, one third of airliners operation costs corresponds to aircrafts fuel. Opticlimb is a software originally developped by the start-up Safety Line (recently aquired by SITA) to compute fuel efficient trajectories and reduce CO2 emissions of aircrafts during their climb. This is achieved thanks to state-of-the-art optimal control techniques and custom statistical models of each aircraft dynamics, learned from historical black-box data. I had the opportunity to contribute to a great extent to this technology during my PhD thesis at the Commands team.