Your goal during this internship will be mainly to iterate over our presentation attack detection solution, to detect frauds based on smartphone screens or silicone masks, for example.

You will focus on:

• Implementing state-of-the-art deep learning algorithms, basing yourself on extensive research of the existing literature and your knowledge of machine learning.

• Optimizing current algorithms, by applying compression techniques to limit the size and inference time of Unissey’s models.

• Developing techniques that mitigate bias to help Unissey building a fairer AI that works equally well for everyone.

Working at Unissey, you may also have the opportunity to contribute to our whole analysis flow, from face detection, to face comparison, deepfake detection, or age estimation.

• Pursuing an MSc / MEng degree in computer science, applied mathematics or any related technical field.

• Prior experience in building and training deep learning algorithms.

• Good theoretical knowledge in computer vision, neural networks, and other DL techniques.

• Excellent coding skills in python (experience with pytorch is a plus).

• Great sense of initiative, autonomy and scientific rigour.

• Curiosity and a strong interest for DL/ML topics and the latest scientific breakthrough.

• Proficiency in English (ability to quickly read and understand research papers).

• Theoretical knowledge of video and image compression techniques is a plus.

• Prior experience with Git and collaborative tools is a plus.

• Personal side projects or Kaggle competitions are a plus.

Details :

• Duration 📈: 6 months.

• Salary 💵: according to profile and past experiences.

• Location 📍: 4 rue du Caire, 75002 Paris.

• Start date 📅: March/April 2025.

• Quick screening call. 📲

• Technical test, to be completed at home. 🪑

• Technical interview with different members of the R&D team. 🌟

• If needed, a short call with a HR employee. 🎀