Smart Interaction and Machine Intelligence Technologies Lab


SiMiT Lab’s objective is to conduct research on computer vision and deep learning, specifically face analysis and other biometric traits within the scope of the international and industrial research projects at the lab.

SiMiT Lab’s research is categorized into three main groups:


Core facial image processing

  • Face detection and recognition
  • Facial feature localization
  • Face modeling
  • Gender and age estimation

Abstract & subjective face analysis

  • Facial expression recognition
  • Action unit detection
  • Action unit intensity estimation
  • Analysis of attractiveness

Mobile gaming

  • Social games
  • Games with computer vision
  • Augmented reality
  • Human computer interaction games


SiMiT Lab is directed by Prof. Dr.-Ing. Hazım Kemal Ekenel, who has over ten years of experience in computer vision and machine learning. Previously, he has founded the Facial Image Processing and Analysis Group in Karlsruhe Institute of Technology (KIT), Germany with a young investigator grant he was awarded within the framework of the Excellence Initiative of German Research Foundation (DFG) in 2010. In 2008, he has received the EBF European Biometric Research Award for his contributions to the field of face recognition during his doctoral studies.


Recent demo videos from the projects/publications

  • A Face Recognition Based Multiplayer Mobile Game Application (AIAI 2014)

  • Cleaning Up After a Face Tracker: False Positive Removal (ICIP 2014)

  • Facial Expression Recognition on a Smartphone

  • 2D-to-3D Facial Expression Transfer (SIU 2016)

  • Real-time Face Swapping in Video Sequences (VISAPP 2012)

  • Automatic Analysis of Facial Attractiveness from Video (ICIP 2014)

  • Real-time Action Unit Intensity Estimation (CVPR4HB 2011)

  • Game Design for Physical Therapy and Rehabilitation Using Kinect (TIPTEKNO 2015)

  • Color Splash Game

  • ITU Virtual Assistant

  • Clothing Matching on a Smartphone

  • Blink Counter: A Plug-in to Monitor Blink Rate(TIPTEKNO 2015)