Computer Vision & Signal Processing

Computer Vision & Signal Processing

Berrin Yanıkoğlu
Berrin Yanıkoğlu
  • Computer Vision & Signal Processing
    A novel biometric privacy framework based on layering multiple biometrics in one template

    Research in Computer Vision and Pattern Recognition areas aim to automatically classify a given image or signal, such as the object in an image. Our groups work on three sub-areas:

    • handwriting recognition for English and Turkish (recognizing on-line and offline handwritten text, arithmetic formulas)
    • biometrics (matching the fingerprint or signature to a reference template)
    • plant identification (recognizing the plant in a given image)

    Our systems have received first place in several international signature verification and plant identification competitions, including SVC2004, 4NSIGCOMP2010, and SigWiComp2013 for online and offline signature verification; and ImageCLEF’2012 for plant identification.

    Further Information
Müjdat Çetin
Müjdat Çetin
Machine Learning-based Methods in Computer Vision and Signal/Image/Video Processing

Müjdat Çetin is affiliated with the VPA and SPIS Laboratories. His research under this theme involves:

  • Facial expression analysis, facial feature tracking, human tracking, and human activity recognition on video data obtained by a single camera or visual sensor networks;
  • Machine learning for sparse and parsimonious signal and image representation;
  • Shape-learning based image segmentation;
  • Machine learning methods for EEG-based brain-computer and brain-machine interfaces;
  • Pattern recognition for remotely sensed data.
Further Information
Associated Faculty Members from other programs are shown in italics