Action intention recognition for proactive human assistance in domestic environments
The current Master’s Thesis in Automatics, Control and Robotics covers the development and implementation of an Action Intention Recognition algorithm for proactive human assistance in domestic environments. The proposed solution is based on the use of data provided by a real time RGBD Object Recogn...
| Author: | |
|---|---|
| Format: | master thesis |
| Publication Date: | 2019 |
| Country: | España |
| Institution: | Universitat Politècnica de Catalunya (UPC) |
| Repository: | UPCommons. Portal del coneixement obert de la UPC |
| Language: | English |
| OAI Identifier: | oai:upcommons.upc.edu:2117/181425 |
| Online Access: | https://hdl.handle.net/2117/181425 |
| Access Level: | Open access |
| Keyword: | Automatic control Robotics Reconeixement de formes (Informàtica) Control automàtic Robòtica Robòtica -- Aspectes socials Àrees temàtiques de la UPC::Informàtica |
| Summary: | The current Master’s Thesis in Automatics, Control and Robotics covers the development and implementation of an Action Intention Recognition algorithm for proactive human assistance in domestic environments. The proposed solution is based on the use of data provided by a real time RGBD Object Recognition process which captures object state changes inside a defined region of interest of the domestic environment setup. A background analysis is performed to analyze state of the art approaches to both real time RGBD object recognition and action intention recognition methods. The preliminary analysis serves as the base for the proposal of a new volume descriptor for object categorization and an improved formalism for Activation Spreading Networks in the context of action intention recognition. Several tests are performed to study the performance of the proposed solution and its results are analyzed to define the conclusions of the project and propose future work. Finally, the project budget and environmental impact as well as the project schedule are presented and briefly discussed |
|---|