Development of a Human Activity Recognition Component
May 28, 2019
Introduction :
Hi, I’m Mariyam. I am a master student at Technical University of Munich, majoring in Computer Science. I like machine learning and robotics.
About the Project:
Understanding actions and interactions of humans from RGB-D sensor input can significantly improve cognitive functions of robots and help safely and smoothly incorporate them in the world of humans. Human activity recognition component will thus be a great addition to the functionality of RoboComp.
Human activity recognition in the video is an interesting and challenging task and recent research shows that there are different ways to address the spatial properties of human actions and their temporal dynamics. In the course of this project I plan to start with a simpler classifier model for CAD-60 dataset and iteratively test and improve the architecture to achieve state-of-the art results on this and other selected datasets with the final goal of providing ready-to-use RoboComp component.
Expected Milestones are:
| Task Description | Phase of GSoC |
|---|---|
| First Iteration of the activity classification development. Dataset analysis and feature extraction, implementation of the appropriate classifier. | Phase 1 |
| Second Iteration of the activity classification development. Experiments with architectures to improve the classifier’s accuracy. Can be specified more precisely after the results of the first phase. | Phase 2 |
| Integration of the HAR component into Robocomp, Testing, Final Documentation. | Phase 3 |
Mariyam Fedoseeva