CYHCDec 31, 2014

iReclass - An automatic system for recording classes

arXiv:1501.00149v11 citations
Originality Synthesis-oriented
AI Analysis

This system addresses the need for automated lecture recording for educators and students, but it is incremental as it builds on existing tracking and gesture recognition technologies.

The paper tackles the problem of automatically recording traditional classroom lectures by developing a system that tracks the lecturer's movements and recognizes five arm gestures to control recording scenarios, achieving a frame rate of 25 fps.

This paper presents the details of a system capable of recording on video a traditional class. By traditional class it is meant a teacher, a blackboard and a white canvas where course notes are projected. The system is able to track the movements of the lecturer, while recording it on video at the required frame rate (e.g., 25 fps). The system is also capable of understanding five arm gestures made by the lecturer with the intent of controlling which scenario is recorded: himself, the blackboard or the white canvas. The remaining two gestures are for start/stop the recorder. The system is composed by a Kinect sensor, a video camera, a microphone, one pan-tilt system and one pan system, using a total of three step motors.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes