Departments |
Zienkiewicz Institute for Modelling, Data and AI
|
Description | Eye tracker provides an intuitive interface to track human eye gaze and fixation. Differ from traditional input (e.g. keyboard, mouse, touch, and speech), these eye movement data are intrinsic, instinctive and immediate response from human. They often reveal the true human's attention, the intention behind the interaction (e.g. lie detection), and is an indispensable data source for human-centric attention research. The goals of this project is 1) to collect eye tracking data from participants using an high end eye tracker (Tobii X3-120 and the underlying APIs), 2) to develop tool to visualize, report and analyse the collected data in term of accuracy and 3) to develop a deep learning technique for automatic analysis and prediction. The implication of the study will offer insights into human perceived saliency on images (e.g. painting, arts, advertisements) and video (e.g. for the development of autonomous car). You will follow existing codes with supports, develop tools, and use mostly python programming (e.g. Tobii Python API and Tensorflow). This project is good to show off your skills for jobs like Data Scientists |
Preparation | As this project uses deep learning, a graphic card will be useful for training. Otherwise, you should consider to use external resource (e.g. Google colab), which may impose time/cost constraints. |
Project Categories |
Architectures/Networks, Artificial Intelligence (AI), Data Science, Visual Computing |
Project Keywords |
Computer Vision, Image Processing, Machine Learning, Neural Networks |
Level of Studies
|
Level 6 (Undergraduate Year 3) |
yes
|
Level 7 (Masters) |
yes
|
Level 8 (PhD) |
yes
|