Research

Ongoing Research:

• Sensory Substitution and Multi-sensory Perception
  Humans and most other animals experience the world through all available sensors as a whole and each sense plays a unique and critical part in fabricating the world we perceived in our mind. (We not only see but also feel, hear, smell and sometimes taste things in our surroundings) There are interesting but yet unexplored relationships among different human sensors, i.e. vision, touch, hearing, smell and taste. Detail study of these sensory perceptions will enrich several other research and technologies.
  • Sensory Substitution
  • Virtual Reality
  • Immersive Environments
  • Simulations
  
  
• Multi-modal Texture Perception
  The word 'texture; is being used widely in several distinct areas, hence its meaning depends on the context it is being used and the sense used to perceive. We are using the word 'texture' to discuss aspects of images/videos (visual), sounds/music (acoustic) and tangible objects/surfaces (haptics). Lack of globally agreed upon definition for texture, specially related to vision and acoustics, makes it a hot tempting research topic. Some of the ongoing research in texture includes
  • Subjective and objective measurement for visual texture roughness
  • Modeling acoustical properties such as roughness, sharpness, timbre etc
  • Visual texture Analysis/Synthesis
  • Limits and dimensions of tactile texture perception
  
  
• Subjective Image Concepts
  Concepts such as 'warm', 'crispy', 'pleasant' etc are very subjective and hard to model. When it comes to image editing however, application of such concepts to a raw image can be of much use. Unfortunately a common user doesn't know a direct mapping from such high-level concepts to available parameters/controls (such as contrast, color, sharpness etc) in image editing software like Gimp or Photoshop. Apart from understanding and modeling the mapping between high-level concepts and low-level image parameters, the research focuses on developing a system that can generate simple controllers to manipulate images with relate these image concepts once it was trained by a user. A detailed description about the class project, which was conducted initially to demonstrate this idea, can be found in this website.

Other Interests

• Telecommunication and Data Communication
• Sensor Networks
• Machine Learning
• Computer Vision
• Computer Graphics
• Financial Engineering
• Culture and Literature