top of page
Active Research Work
Research-Problem
Research-Team
Single Image Dehazing
The images captured in poor weather degrade due to atmospheric scattering, resulting in poor performance of outdoor vision based systems such as surveillance, object detection, object tracking, etc. The dehazing is a process to obtain a haze-free image from a single hazy image.
Paras Garg
UG Student
Under-water Image Enhancement
The images captured inside the water degrade due to scattering, and absorption of light by water particles. This reduces visibility inside the water. Hence, the exploration of sea is difficult. The aim of under-water image enhancement is to improve visibility of the images captured inside the sea world.
Gunjan Vrma
Ph.D. Candidate
Input
Output
Single Image Super-Resolution
The images captured with poor quality of camera or from the long distance, blurred due to low resolution. The super-resolution process increases the resolution of the images by interpolation. The super-resolution takes a low resolution image as input and produces respective high resolution image as output.
[OUTCOME: A paper titled "Structure Preserving Loss Function for Single Image Superresolution" has accepted for publication in the Journal of Displays (IF:3.074, SCI Indexed, Elsevier). ]
Naman Tuli
UG Student
Image Captioning
The image captioning is a process to describe an image through texts. This process takes an image as input and returns the description of the images by a set of words. Image Captioning is an important area of research for automatic description of a scene.
Shreya Garg & Team
UG Student
Face Unmasking
The peoples wear mask under poor atmospheric conditions and in health issues. Hence, the vision systems based on face data perform poorly. This necessitates automatic unmasking of the face. The unmasking takes a masked face image as input and produces unmasked face image as output. [In collaboration with Research and Development Wing of Samsung, India]
Kaustuv
UG Student
bottom of page