Thema
Video Saliency Metrics Implementation
Typ Master, Forschungspraxis, Bachelor, Ing.prax.
Betreuer Mikhail Startsev
Tel.: +49 (0)89 289-28550
E-Mail: mikhail.startsev@tum.de
Sachgebiet Human and Computer Vision
Beschreibung One of the important questions in computer vision is how you determine what information in a scene is relevant. So-called “saliency models” [1] have been used to predict informativeness in images [2] and videos. A quality ground truth labelling of the eye movements is what provides for an accurate and versatile evaluation of saliency models. This labelling is however, especially for video sequences, hard to come by.

Our group has been working on obtaining such ground truth data. And while this effort is still ongoing, evaluation of the existing saliency models can already begin. The candidate would implement various metrics that describe just how well the predicted salient regions fit the actual distribution, described by ground truth. This would help verify our implementations of the same metrics as well as broaden the spectrum of the evaluation procedure.


Image source: jov.arvojournals.org/article.aspx

[1] en.wikipedia.org/wiki/Salience_(neuroscience)
[2] saliency.mit.edu/results_mit300.html
Voraussetzung Knowledge of machine learning concepts and solid programming skills are required.
Bewerbung If you are interested in this topic, we welcome the applications via the email address above. Please set the email subject to “<Type of application> application for topic 'XYZ'”, ex. “Master’s thesis application for topic 'XYZ'”, while clearly specifying why are you interested in the topic in the text of the message. Also make sure to attach your most recent CV (if you have one) and grade report.