InFrame

Project Introduction

Robotic-assisted photography is among the most exciting technological developments in the production industry since the invention of the camera itself. It provides new and unique ways to capture dynamic scenes and unlock new avenues for cinematographic creative potential. Current photography robots on the market range from tremendously expensive industrial robotic arms, such as Motorized Precision’s Kira camera robot, to portable devices that allow for smooth motion across scenes like the Rhino Arc II. However, most photography robots have not yet leveraged breakthroughs made in the Computer Vision space in the last decade. For these reasons, we created InFrame: an intelligent, motorized photography assistant that uses state of the art object detection models and tracking algorithms to follow user-selected targets across a 3D space, constantly keeping them InFrame.

InFrame was built by Diego Martinez, Ike Kilinc and Ismael Mercier as part of our S20 Senior Capstone Project for Electrical and Computer Engineering at Carnegie Mellon University. The code and CAD files for InFrame can be found in our Github Repository.