I'm a Master’s student in Computational Sciences and Engineering at Harvard University. I hold a joint research appointment - a Research Assistantship at MIT and Harvard SEAS Fellowship. I am co-advised by Fredo Durand and Aude Oliva at MIT, and Hanspeter Pfister at Harvard. My current work focusses on applying tools from computer vision and human perception to automatic understanding of multi-modal images like Infogrpahics. Here is our most recent paper on this.
Before this, I completed my Bachelor and Master of Technology from the Indian Institute of Technology Delhi (IIT Delhi) in 2015. In the past, I have also worked as a full time research engineer at the University of Pennsylvania on developing metrics for comparing unsupervised clustering algorithms.
I am extremely passionate about teaching, and in my free time I work on Machine Learning tutorials, give lectures/talks on computer vision and maintain a blog which is loosely connected to science and education. Find our more below!
Madan, S.*, Bylinskii, Z.*, Tancik,M.*, Zhong, K., Recasens, A.,Alsheikh, S., Pfister, H., Durand, F., 2017. Iconness: an Icon Proposal Mechanism for Annotating and Summarizing Infographics. [in submission]
Bylinskii, Z.*, Alsheikh, S.*, Madan, S.*, Recasens, A.*, Zhong, K., Pfister, H., Durand, F. and Oliva, A., 2017. Understanding Infographics through Textual and Visual Tag Prediction. arXiv preprint arXiv:1709.09215.[Link]
Zoya Bylinskii, Nam Wook Kim, Peter O'Donovan, Sami Alsheikh,Madan, S., Hanspeter Pfister, Fredo Durand, Bryan Russell, Aaron Hertzmann. Learning Visual Importance for Graphic Designs and Data Visualizations. UIST 2017. [Link]
Shayoni Dutta , Spandan Madan , Harsh Parikh, Durai Sundar. An ensemble micro neural network approach for elucidating interactions between zinc finger proteins and their target DNA. BMC Genomics, 17(13), 97 (2016). [Link]
Shayoni Dutta ,Spandan Madan, Durai Sundar. Exploiting the recognition code for elucidating the mechanism of zinc finger protein-DNA interactions. BMC Genomics, 17(13), 109. (2016). [Link]
Gave a talk at Harvard Business School on an overview of successes in Deep Learning.[Link]
Gave a talk at HackMIT on AI v/s Deep Learning v/s Machine Learning.
Gave a talk at MIT Blueprints on Computer Vision.[Link]
My Tutorial featuring an end to end M.L. pipeline - data collection, problem framing, implementation. [Link]
The problem with math is that it takes a long, beautiful road to create every concept in your book, but sadly, once we're there the path is forgotten, and a "Definition" is coined. To teach math is therefore no other than storytelling - the story of why it was needed, and how it was used"
- Brad Osgood, Stanford EE261 lectures
This blog is an attempt to present my views on science and how scientific education should be approached!Visit the blog
Music is a big part of my life and here you can have a glimpse at some of the songs I love listening to, and the covers/originals I have recorded in the past!