Faculty of Engineering and Computing - Research

Faculty of Engineering and Computing

Congratulations to Neil O'Hare

 

 

17 September 2007

Neil O'Hare - PhotoCongratulations to Neil O'Hare who successfully defended his thesis and will be awarded the degree of PhD.

 

The title of Neil's thesis is "Semi-Automatic Person-Annotation in Context-Aware Personal Photo Collections".

 

He completed his PhD in the Centre for Digital Video Processing (CDVP), Adaptive Information Cluster (AIC) and the School of Computing, DCU under the supervision of Prof. Alan Smeaton.

 

Brief description of Project:

The MediAssist system is a prototype context-aware photo management system that facilitates browsing, searching and semi-automatic annotation of CDVP - Logopersonal photos. Within this system, we developed an approach to semi-automatic person-annotation in personal photo collections that facilitates the annotation of people in a batch manner by suggesting person names for photos as users interact with the system.

 

We developed person classification and retrieval techniques based on analysis of the context of photo capture, such as the time and location, in addition AIC - Logoto analysis of the image content of the photo. We used classification techniques to suggest names for faces detected in photos, and retrieval techniques to suggest faces for a query name. We implemented the proposed techniques and integrated them into the interface of the MediAssist prototype photo management system.

 

We successfully evaluated our person classification and retrieval techniques using the real photo collections of a number of users, and we also successfully evaluated the interactive annotation system with a number of users.

 

Object tracking, a complementary technique to object detection, was also explored in a multi-source context and an efficient framework for robust tracking, termed the Spatiogram Bank tracker, was proposed as a means to overcome the difficulties of traditional histogram-based tracking. While general frameworks for data fusion were developed, the fusion of thermal infrared video with standard visual video was specifically targeted. Potential applications include: pervasive human-computer interaction and automated surveillance.

 

Enterprise Ireland - LogoThis project was generously funded by Enterprise Ireland.

 

 

 

 

 

Back to News Headlines