Harish Katti

Hi, I m Harish and am doing post-doctoral research in the vision research group at the Center for Neuroscience, Indian Institute of Science. My Interests are in the intersection of human cognition and visual media. I have bachelors in computer science from Karnatak University, masters in bio-medical engineering from IIT Bombay and a PhD from the National University of Singapore. More about me.



Post-doctoral research: Computational mechanisms of visual object categorization, Mentor: Asst Prof S. P. Arun

Our everyday visual experience frequently involves finding objects such as a face in a crowd. Such rapid visual categorization involves delineating objects from background (segmentation), and matching their features with those of the target (recognition) and making sense of the overall scene (context). Damage to specific brain areas can lead to difficulties in segmenting objects (simultagnosia) or recognizing objects (e.g. visual agnosia). How does the brain accomplish segmentation and recognition? Do these two processes interact and what are their neural mechanisms? Despite many advances in our understanding of visual representations in the brain, we do not yet understand how segmentation interacts with recognition. Computational investigations of this problem have shown that delineating objects from background and their subsequent recognition need not be distinct processes. The goal of the proposed research is to investigate the interplay between segmentation and recognition in the human brain. I propose to use state-of-the-art computer vision algorithms and behavioral experiments in humans to elucidate the computational mechanisms of visual object segmentation, recognition and context processing.

PhD Thesis: Human Visual Perception, study and applications to understanding Images and Videos, PhD Supervisors: Prof. Mohan Kankanhalli, Prof. Tat Seng Chua, Panel Members: Asst. Prof Terence Sim, Prof Tan Chew Lim

Specific aims addressed in PhD thesis:

Aim 1: The important role of color in aesthetics discrimination in very short presentation times and also established that humans can distinguish simple notions of aesthetics even at pre-attentive time spans (<100ms).

Aim 2: Influence of visual cues relating to objects, interactions and affect on eye movements. A model that quantifies attentional bias amongst important concepts in social scenes is developed along with a novel clustering algorithm and measures to infer interaction and location of visual concepts in images using eye-movements.

Aim 3: Applications including detection of key objects and interactions, foreground object segmentation and localization of text keywords into image content. 

Multimedia Analysis and Synthesis Lab ( photo of teachers day 2009 ), Department of Computer Science, School of Computing,

News and Updates

I have joined the Center for Neuroscience, Indian Institute of Science as a post-doc in the Vision research group( S P Arun lab).

Thesis defense done, (May 2 nd, June 2012).

I submitted my thesis for review!!  I am looking for post-doc positions in computational/experimental vision research and a brief description of my research interests is here (pdf).

Our bid for a tutorial on eye-tracking methodology has been accepted at ACM Multimedia 2011 !!

NEW !! Updated copyright information and up to date mirror for the NUSEF dataset

My Work was part of the Panel discussion on the Use of Non-Conventional means for Media Content and analysis at ACM Multimedia 2010. Prof Mohan was one of the panelists . (slides)


Congyan Lang*, Harish Katti*, Tam Nguyen*, Karthik N. Yadati, Depth Matters: Influence of depth cues on visual saliency, Accepted in the 12th European conference on Computer Vision (ECCV ‘12), Florence, Italy. * indicates equal contribution.

Harish Katti, Karthik Yadati, Mohan Kankanhalli, Chua Tat-Seng, Affective video summarization and story board generation using Pupillary dilation and Eye gaze, The IEEE International Symposium on Multimedia 2011 (ISM 2011), December 5-7, 2011, Irvine, USA bibtex, pdf, Oral presentation slides 

Harish Katti, Mohan Kankanhalli, Eye-tracking methodology and applications to images and video, In Proceedings of the international conference on Multimedia (MM ‘11). ACM, Scottsdale, Arizona, USA, Nov 28-Dec 1, 2011, (ACM digital library link), bibtex, pdf, tutorial slides

Harish Katti, Ramanathan Subramanian, Mohan Kankanhalli, Nicu Sebe, Tat-Seng Chua, and Kalpathi R. Ramakrishnan. 2010. Making computers look the way we look: exploiting visual attention for image understanding. In Proceedings of the international conference on Multimedia (MM ‘10). ACM, New York, NY, USA (ACM Digital library link), bibtex, pdf

Subramanian Ramanathan *, Harish Katti *, Nicu Sebe, Mohan Kankanhalli, and Tat-Seng Chua. 2010. An eye fixation database for saliency detection in images. In Proceedings of the 11th European conference on Computer vision: Part IV (ECCV 10), Kostas Daniilidis, Petros Maragos, and Nikos Paragios (Eds.). Springer-Verlag, Berlin, Heidelberg, 30-43. (ACM Digital Library Link), bibtex, pdf, NUSEF dataset link. (* indicates equal authorship) … Our paper on the NUSEF dataset

Subramanian Ramanathan *, Harish Katti *, Raymond Huang, Tat-Seng Chua, and Mohan Kankanhalli. 2009. Automated localization of affective objects and actions in images via caption text-cum-eye gaze analysis. In Proceedings of the 17th ACM international conference on Multimedia (MM ‘09). ACM, New York, NY, USA, 729-732. (ACM Digital Library link), bibtex, pdf . (* indicates equal authorship)

Katti, H.; Kwok Yang Bin; Tat Seng Chua; Kankanhalli, M.; , "Pre-attentive discrimination of interestingness in images," Multimedia and Expo, 2008 IEEE International Conference on , vol., no., pp.1433-1436, June 23 2008-April 26 2008 , (IEEE Explore link) , bibtex, pdf, Oral presentation slides

Other manuscripts

Thesis proposal presented on 2 nd March 2011. Slides from the presentation.

Graduate Research Paper on Multimedia Semantics, 7th Nov, 2007 (pdf). The accompanying presentation slides (Panel was Prof. Terence Sim and Prof. Tan Chew Lim). Passed !

A report on “RTP compliant MP3 Audio Packetization for the Yima PE streaming system” for Prof. Zimmerman s course on streaming media (13th Nov, 2007)

ACM SIG CHI 2009 A non-invasive framework for interest and affect modeling in images (video, poster at http://www.comp.nus.edu.sg/~harishk/chi2009/) this one didn t make it :P

Generative models in multimedia content

Presentations and Talks

Eye-gaze and image semantics at CVAI lab, IISc and Center for Neuroscience, IISc and Yahoo Research machine learning summer school 2010, Bangalore (ppt).

Media semantics in the Max Planck Institute for Cybernetics, Tuebingen .

Image interestingness for the CANALAVIST talks http://www.canalavist.org/ict-forum/abstractimages.html the presentation slides

Earlier work

Semantics of location (Project TACREA) and inferring locations semantics from web content

Investigating interestingness in images

Generalised imaging

School of Computing
National University of Singapore
AS6, level 5, Room 01
Computing Drive
Singapore 117590

Mobile : (65) 9733-4666 (late evening is a good time to call)
Phone : (65) 6516-4362 (lab)
Fax :
(65) 6777-9096

There is treasure everywhere!! ... Calvin & Hobbes