Prophesee is the inventor of the world’s most advanced neuromorphic vision systems.
The company developed a breakthrough Event-Based Vision approach to machine vision. This new vision category allows for significant reductions of power, latency and data processing requirements to reveal what was invisible to traditional frame-based sensors until now.
Prophesee’s patented Metavision® sensors and algorithms mimic how the human eye and brain work to dramatically improve efficiency in areas such as autonomous vehicles, industrial automation, IoT, security and surveillance, and AR/VR. More information can be found at www.prophesee.ai.

Event-based cameras represent a revolutionary approach to capturing motion, offering high temporal resolution, low latency, and a unique data representation that is ideal for dynamic scenes.

Building on previous work in 3D body pose estimation, this internship focuses on extending these techniques to the domain of gesture recognition, specifically inferring the 3D pose of the hands using monocular event-based cameras.
To achieve this, we will use active markers—LEDs that transmit data through variable frequency blinking—as ground truth. The project involves building gloves equipped with these active markers and using a system of cameras to triangulate their positions, providing accurate 3D data for training machine learning models. This work is a critical step toward developing real-time, accurate gesture recognition systems with applications in VR/AR, human-computer interaction, and more.

Key Responsibilities

• 3D Hand Pose Estimation: Develop and train machine learning models to infer 3D hand poses from raw event data captured by event-based cameras
• Active Marker System Development: Build gloves equipped with active IR LEDs, ensuring they are optimally designed for reliable tracking (this may involve electronics assembly, soldering, and microcontroller programming.)
• System Calibration: Calibrate the intrinsic and extrinsic parameters of the camera system to ensure accurate 3D triangulation and reprojection
• Microcontroller Programming: Work with microcontrollers (e.g., Raspberry Pi Pico) to control the active markers (this may include diving into assembly code to optimize performance and ensure precise timing of the LED signals)
• Data Acquisition and Ground Truth Collection: Design and execute a robust data acquisition campaign using the active marker-equipped gloves to gather ground truth data for training and validating the models
• Experimentation and Analysis: Conduct experiments to validate the models, analyze the results, and iterate on model designs to achieve optimal performance
• Documentation: Document the methodology, code, and findings thoroughly to ensure reproducibility and future development continuity
  

Your profile

• Master Degree
  

• Programming skills in Python, C++, or similar
• Experience with machine learning frameworks (e.g., PyTorch, TensorFlow)
• Basic knowledge of electronics, including soldering and microcontroller programming
• Familiarity with 3D geometry, camera calibration, and pose estimation techniques
• Knowledge of event-based camera technology and active marker systems is a plus

• Analytical Mindset: Ability to analyze complex problems, conduct experiments, and interpret results
• Communication Skills: Written and verbal communication skills for documentation and team collaboration
• Passion for Innovation: Eagerness to explore new technologies and push the limits of current methodologies

Our Benefits

Salary package : 80% SMIC (9,32 €/h)
Meal vouchers : 9€ / working day (Swile card)
Working hours : 35h / week
Office location : Paris Bastille (74 rue du Faubourg Saint Antoine 75012)
Other benefits : CSE, Incluson & Diversity

We believe in fostering an inclusive environment where everyone is valued and respected. We welcome candidates from diverse backgrounds and are committed to ensuring equal opportunities for all, regardless of race, gender, disability, or any other characteristic.