MSc graduation project: an intelligent agent that learns to read text

Context

BIASlab logo

BIASlab (Fig.1, http://biaslab.org, FLUX-7.060) is a subgroup of the Signal Processing Systems (SPS) that aims to develop Intelligent Autonomous Agents (IA). These agents interact with their environment through their sensors and actuators in order to learn purposeful behavior, e.g., to navigate, play soccer or they may learn to decode speech signals under bad acoustic conditions. Our research projects are inspired by the latest insights from machine learning, computational neuroscience and signal processing.

Project Description

hierachical-brain-processing

In this project you are challenged to develop a software demonstrator for an intelligent agent that learns how to read text. Text reading by humans does not proceed by scanning character sequences in the same order as they appear in the text. Rather, the look direction for the eyes is continually updated to maximize information gain, i.e., to maximally resolve uncertainty about the textual contents. This kind of reading process requires intelligent dynamical hierarchical reasoning at various levels (characters, words, sentences, etc.), see Figure (from: Friston K., Hierarchical Models in the Brain, PLoS Comput Biol 4(11), 2008). In this project, we combine insights from diverse fields such as computational neuroscience, machine learning and signal processing and develop a software demonstrator for an intelligent reading agent. The agent senses images of characters and controls the look direction for the eyes to scan for the next character. The project is ideally suited for a student who wants to learn about neuroscience-inspired machine learning techniques. You will get ample support and time to develop your knowledge of biological intelligent systems and how to translate those insights to engineering systems.

Timing

Start date: the project is available from April 2018 (or any time thereafter).

Duration: 9 months (fte).

Financial Support

The audio solutions company GN (http://www.gn.com/) may financially support strong candidates (qualifications to be discussed with prof. de Vries) by a GN scholarship.

Contact

For more information about this project, please contact Prof. Bert de Vries (bert.de.vries@tue.nl). Also, feel free to make an appointment to discuss alternative projects with intelligent autonomous agents that you might be interested in.