Computer Vision Engineer – Intern

About FuturHand Robotics

FuturHand Robotics is a Carnegie Mellon University spin-off developing next-generation dexterous robotic systems that transform how robots interact with the world. We build on cutting-edge research in dexterous hands and manipulation to solve real-world problems, developing dexterous bimanual robots that operate reliably outside controlled lab environments.

The Role

We're looking for a Computer Vision Intern to join our perception team and work directly with our Computer Vision Lead. You'll get hands-on experience with real perception problems on a physical robotic platform — not synthetic benchmarks or toy datasets. This is a great opportunity to see how vision systems are built, tested, and deployed in a production robotics stack, and to contribute meaningfully to a small, fast-moving team.

You'll take ownership of a well-scoped project or set of tasks within the broader perception pipeline spanning sensor data collection & processing, object detection, pose estimation, or scene understanding with guidance and mentorship from the CV Lead. As you ramp up, you'll have room to take on more open-ended problems.

What You'll Be Doing

• Support development of components of the perception pipeline, such as sensor integration, object detection, pose estimation, or state tracking
• Help build and improve tooling for perception: data pipelines, evaluation scripts, debugging/visualization tools, and performance metrics
• Run experiments to test robustness across variation in lighting, object geometry, and scene complexity
• Collaborate with the CV Lead and broader manipulation/controls teams to understand how perception outputs are used downstream
• Explore techniques such as sensor fusion, temporal filtering, or data-driven methods to improve system robustness
• Document findings and contribute to the team's understanding of what works and what doesn't

What We're Looking For

• Currently pursuing a BS (upper-class/senior) or MS/PhD in computer vision, robotics, computer science, or a related field
• Solid fundamentals in computer vision with coursework or project experience with 3D perception, object detection, or pose estimation
• Comfortable programming in Python (and/or C++); familiarity with common CV/ML frameworks (e.g., PyTorch, OpenCV)
• Exposure to depth sensors, RGB-D data, or 3D geometry is a plus, but not required
• Prior experience working with camera hardware and camera calibration is a plus, though not required
• Familiarity with data collection, data cleaning and data labelling
• Curiosity about how perception fits into a larger robotic system, and willingness to learn how design choices affect downstream performance
• Based in or willing to relocate to Pittsburgh, PA for the duration of the internship

Nice To Have

• Prior project or coursework experience with robotics perception for manipulation tasks
• Familiarity with ROS 2
• Exposure to sim-to-real transfer, domain randomization, or data augmentation
• Experience deploying models on embedded or edge hardware (e.g., Jetson)

What We Offer

• Direct mentorship from our Computer Vision Lead and close collaboration with a small, highly technical team
• Real ownership of a project that contributes to a novel robotic platform, not busywork
• Exposure to the full perception stack — from sensors to deployed, real-world behavior
• Competitive internship compensation
• Potential path to a full-time offer for strong performers

How to Apply

Email us with your resume and a short note describing a vision system you've built. What did the system need to handle, and what were the hardest parts to get right?

[email protected]