Research Interests

• Machine Learning\Pattern Recognition
• Semi-Supervised Methods
• Dimension Reduction Methods
• Gaussian Processes
• Human Activity Recognition
• 3D Human Pose Estimation
• Computer Vision

Research Experiences

• Digital Media Lab (DML), Advanced Information and Communication Technology Center (AICTC) of Sharif University. 2011-2013

  3D human pose estimation
  We presented a semi supervised graph based method that estimates 3D body poses from a sequence of silhouettes. Due to ambiguity in the silhouettes, shortcut edges are created through the constructed graph process. In order to increase the performance of estimation and also to reduce MSE (Mean Square Error), we introduced a approach for removing of these shortcut edges.

• Digital Media Lab (DML), 2010

  Community Detection in Social Networks
  We studied various global community detection approaches which discover all communities in a graph based on topological information of the graph.

• Digital Media Lab (DML), 2009

  Applications of Game Theory in Wireless Ad-hoc Networks.
  Using game theory to energy management in wireless ad-hoc networks.

Publications

• Mahdieh Abbasi, 3D Human Pose Estimation, Master thesis, Alzahra University, 2012.


• M. Abbasi, H. R. Rabiee, R. Azmi, Monocular 3D Human Pose Estimation with a Semi-supervised graph based method, submitted in Pattern Recognition Letters, 2014.