Machine Learning Engineer - Robot Perception

7 months ago

H1B Sponsor

Responsibilities

Develop, train, and deploy machine-learning perception algorithms for object detection, pose estimation, tracking, and scene understanding.
Integrate sensor-fusion techniques using cameras, depth sensors, IMUs, and tactile feedback.
Optimize real-time perception pipelines for low latency and robust performance in dynamic environments.
Work with hardware engineers to design sensor configurations and optimize perception models for onboard deployment.
Integrate perception systems with reasoning, manipulation, planning, and control capabilities.
Collaborate across disciplines and provide technical mentorship to junior engineers.

An MS or PhD in machine learning, computer science, robotics, or a related field is required.
Strong background in computer vision, deep learning, and sensor fusion is required.
Proficiency in Python and C++, with experience using PyTorch, TensorFlow, OpenCV, and ROS, is required.
Hands-on experience with real-world robotics perception systems such as SLAM, 3D reconstruction, or multimodal perception is required.
Experience setting up and calibrating cameras, LiDAR, and other sensors is required.
Experience with data collection, preprocessing, and management for training machine learning models is required.
Preferred qualifications include familiarity with Gazebo or MuJoCo and experience with sim-to-real transfer.
Preferred experience includes handling noisy, incomplete, or sparse data; deploying models to edge devices such as NVIDIA Jetson; and accelerating training with GPU, TPU, or other hardware accelerators.
General knowledge of robotics principles including kinematics, dynamics, and control is preferred.
The role requires problem identification and execution skills, a self-starter attitude, and enthusiasm for a fast-paced startup environment.

C++OpenCVPython PyTorchTensorFlow

11-50 employees

Maven Robotics is building the world’s leading general-purpose AI robots.