2020-11-06T14:28:31 n8909

Noisy MOBIO Landmarks

Viewed: 3536

Noisy MOBIO Landmarks


Face landmarks for the MOBIO database (https://www.idiap.ch/dataset/mobio) with different level of noise provided to evaluate face recognition in the presence of localisation noise. Contact Dr Simon Denman for further information.


The Noisy MOBIO Landmarks are © 2014 QUT and is licensed under the Creative Commons Attribution-ShareAlike 3.0 Australia License.


To attribute this database, include the following citation: 
K. Anantharajah, Z. Ge, C. McCool, S. Denman, C. Fookes, P. Corke, D. Tjondronegoro, S. Sridharan (2014) Local Inter-Session Variability Modelling for Object Classification. In IEEE Winter Conference on Applications of Computer Vision (WACV) Please note that authors should also cite and acknowledge the MOBIO database as outlined on the MOBIO website.

Downloading and using the Noisy MOBIO Landmarks database

Four sets of landmarks are provided, corresponding to added uniform noise equal to 2, 5, 10 and 20% of the average inter-eye distance:

Each file contains a list of images and new landmark points. Each line consists of (in order) the filename, right eye X coordinate, right eye Y coordinate, left eye X coordinate, and left eye Y coordinate. These landmark files should be used in place of the landmarks provided with the MOBIO database.

Geographical area of data collection



Anantharajah, Kaneswaran, Ge, ZongYuan, McCool, Christopher, Denman, Simon, Fookes, Clinton B., Corke, Peter, Tjondronegoro, Dian W., & Sridharan, Sridha (2014) Local inter-session variability modelling for object classification. In IEEE Winter Conference on Applications of Computer Vision (WACV 2014), 24-26 March 2014, Steamboat Springs, CO. http://eprints.qut.edu.au/67786/

Research areas

Face Recognition; Biometrics

Cite this collection

Denman,Simon. (2016): Noisy MOBIO Landmarks. [Queensland University of Technology]. https://doi.org/10.4225/09/586f297390d99

Data file types

text files


Creative Commons Attribution-ShareAlike 3.0 Australia License.


© Queensland University of Technology, 2014.


Has chief investigator
Simon Denman  (Researcher)


Name: Dr Simon Denman
Phone: +61731389329


Date record created:
Date record modified:
Record status:
Published - Open Access