Computer Vision Scientist

Job Description

At Postmates X , we are reimagining how things move in cities. Serve , our personable sidewalk delivery robot, is our vision for the future of delivery. It’s designed to take deliveries away from congested streets and onto underutilized alleys and sidewalks, to save energy, reduce emissions, reduce traffic, and make deliveries faster and cheaper. Why move two-pound burritos with two-ton cars?

Serve has been years in the making and has already done thousands of customer deliveries.

WHO ARE WE?

We are tech industry veterans in software, hardware and design who are pooling our individual talents together to build a legacy. Our team is inclusive, agile, and driven by a common purpose. We are solving real-world problems with machine learning, computer vision, and robotics, with a mindful eye towards the end-to-end user experience. By joining our team, you will help create the first robotic helpers entering our cities: moving alongside us and running our errands!

WHAT WILL YOU DO?

You will collaborate with a team of multidisciplinary engineers to tackle hard problems and help advance the future of autonomous robots. More specifically, you will apply computer vision and computational geometry fundamentals to real world problems related to autonomous vehicles’ ability to perceive the world around them. As an example, you will work with both 2D and 3D object detection and segmentation using deep learning to facilitate the autonomous robot’s operations. The position consists of the full development cycle where you’d design a test setup for emulating the real-world conditions, develop algorithms to perform the required measurements, test it on the field and refine the product by learning and applying new algorithms.

RESPONSIBILITIES:

• Development of computer vision algorithms that are efficient and robust.
• Working with other software and mechatronics engineers to integrate algorithms into a complete product
• Develop algorithms and systems for stereo and 3D imaging, feature extraction, and sensor fusion
• Build tools and framework for multiple cameras and sensor calibration
• Evaluate new camera models for both 2D and 3D
• Apply machine learning to computer vision problems for segmentation and object detection

QUALIFICATIONS

• MS or PhD degree with 5 years experience in the computer vision field.
• Proficient programming skills in C++ or Python
• Strong knowledge in multiple geometries in computer vision
• Experience in stereo vision algorithms and 3D sensor data (time of flight, structured light, lidars)
• Experience with OpenCV or PCL
• Experience with machine learning algorithms related to segmentation, classification, detection and tracking

BONUS

• Experience with human pose estimation
• Experience with deep learning algorithms: CNN, RNN, GAN and other Generative Models
• Experience with a Deep Learning framework: TensorFlow or PyTorch

BENEFITS

• Competitive salary and generous stock option plan
• Medical, dental and vision insurance
• We'll provide equipment you need to work efficiently and creatively
• Paid parental leave, vacation time and sick time
• Catered lunches and open snack bar
• Impact-first work environment (no politics, no pandering)
• Huge company vision (we need you to build the future, not just maintain the status quo)
• Full support to contribute to open source projects
• Awesome office located in Financial District just minutes from BART, Muni, AC Transit, and SamTrans