Masoud Mazloom

     PostDoc researcher in automated analysis of visual and textual data in Social Media
     ISIS , Informatics Institute, University of Amsterdam

     E-mail:        ,
     Address:               Science Park 904, 1098 XH , Amsterdam, The Netherlands
     Room number:     C3.253
     Phone:                 +31 (0)20 - 525 7588




My research interest focus on the topics in:

  Computer Vision

  Machine Learning

  Deep learning

  Multimedia Analysis

  Video and Image caption generation

My current research focus on Brand Centrality, Emotions and Virality in Social Media where the goal is to analyzing social media from marketing aspect. In this project I advised by Prof. Marcel Worring and Prof. Willemijn van Dolen.

The topic of my research during my PhD was constructing a precise visual event recognition system. I aim to develop a precise and efficient video search engine able to retrieve specific events involving people interacting with other people, objects and animal in different scene such as feeding animal, changing a vehicle tire, wedding ceremony. For this reason we proposed to use semantic features for analyzing an event in a video to automatically interpret what is happening where in video. We show the capability of using semantic features specially on the more realistic scenario where only a few or zero-example for an event are available. My supervisors in this project were Dr. Cees G. M. Snoek and Prof. Arnold Smeulders at Informatics Institutes at University of Amsterdam .

I received a master's degree after graduating from the Sharif University of Technology, where I conducted research at the Image Processing Lab under supervision of Prof. Shohreh Kasaei. In my research I aim at constructing an Intelligent system to automatically identfy people from their face in image dataset. We proposed to use the combination of PCA and Wavelet transform for feature dimension reduction in a face recognition system.




    About me




Curriculum Vitae