Saikat Basu

Research Scientist,
Facebook Maps,
Cambridge, USA.

Google Scholar   DBLP


Research

My area of interest includes Computer Vision and Machine Learning with a focus on Deep Learning algorithms applied to imagery data and I have also worked on areas like Statistical Modeling, Cryptography and Mobile Computing in the past.


In the News

1)  Deep Belief Networks at Heart of NASA Image Classification, The Next Platform.

2)  NASA Using Deep Belief Networks for Image Classification, Nvidia Developer News.

3)  Quantum computer learns to ‘see’ trees, Science Magazine

4)  Quantum computer learns how to 'see' trees, The Inquirer,

5)  Dwave 2X quantum annealing machine 90% accurate in recognizing tree cover in aerial photos which was better than conventional computers, Next Big Future News

6)  DeepSat: Monitoring the Earth’s Vitals with AI, Inside HPC

7)  How HPC is underpinning amazing advances in land, sea and space observation, Verne Global

8)  An AI Monitor of Earth's Vitals, NVIDIA Success Story

9)  Quantum Computer Recognizes Trees with 90% Accuracy, NewsBeat Science Tech


Selected publications

1) Saikat Basu, Manohar Karki, Robert DiBiano, Supratik Mukhopadhyay, Sangram Ganguly, Ramakrishna Nemani, Shreekant Gayaka, A Theoretical Analysis of Deep Neural Networks for Texture Classification, International Joint Conference on Neural Networks, IJCNN 2016. (pdf)

2) Saikat Basu, Manohar Karki, Sangram Ganguly, Robert DiBiano, Supratik Mukhopadhyay, Ramakrishna Nemani, Shreekant Gayaka, Rajgopal Kannan, Learning Sparse Feature Representations using Probabilistic Quadtrees and Deep Belief Nets, Neural Processing Letters. (pdf)

3) Basu, S.; Ganguly, S.; Nemani, R.R.; Mukhopadhyay, S.; Gong Zhang; Milesi, C.; Michaelis, A.; Votava, P.; Dubayah, R.; Duncanson, L.; Cook, B.; Yifan Yu; Saatchi, S.; DiBiano, R.; Karki, M.; Boyda, E.; Kumar, U.; Shuang Li, A Semiautomated Probabilistic Framework for Tree-Cover Delineation From 1-m NAIP Imagery Using a High-Performance Computing Architecture, in Geoscience and Remote Sensing, IEEE Transactions on , vol.53, no.10, pp.5690-5708, Oct. 2015 doi: 10.1109/TGRS.2015.2428197. (pdf)

4) Saikat Basu, Manohar Karki, Sangram Ganguly, Robert DiBiano, Supratik Mukhopadhyay, Ramakrishna Nemani, Learning Sparse Feature Representations using Probabilistic Quadtrees and Deep Belief Nets, European Symposium on Artificial Neural Networks, ESANN 2015. (pdf)

5) Basu S., Karki M., Stagg M., DiBiano R., Ganguly S. and Mukhopadhyay S. (2015). MAPTrack - A Probabilistic Real Time Tracking Framework by Integrating Motion, Appearance and Position Models. In Proceedings of the 10th International Conference on Computer Vision Theory and Applications, ISBN 978-989-758-091-8, pages 567-574. DOI: 10.5220/0005309805670574 (pdf)

6) Saikat Basu, Sangram Ganguly, Supratik Mukhopadhyay, Robert Dibiano, Manohar Karki and Ramakrishna Nemani, DeepSat - A Learning framework for Satellite Imagery, ACM SIGSPATIAL 2015. (pdf)

7) Saikat Basu, Robert Di Biano, Manohar Karki, Malcom Stagg, Jerry Weltman, Supratik Mukhopadhyay, Sangram Ganguly, An Agile Framework for Real-time Motion Tracking, COMPSAC 2015 International Workshop on Consumer Devices and Systems. (pdf)

8) Human Action Recognition in Surveillance Videos using Abductive Reasoning on Linear Temporal Logic, United States Department of Defense (DoD) Technical Report, August 2012.

9) An Agile Framework for Real-Time Visual Tracking in Videos, United States Department of Defense (DoD) Technical Report, September 2012.

10) Quantum Boosting and Fast Classical Metrics for Tree Cover Detection in Remote Sensing Data, AGU Fall Meeting, 2014.

11) Deriving Continuous Fields of Tree Cover at 1-m over the Continental United States From the National Agriculture Imagery Program (NAIP) Imagery to Reduce Uncertainties in Forest Carbon Stock Estimation, AGU Fall Meeting, 2013.

12) A new parallel window-based implementation of the Elliptic curve point multiplication in multi-core architectures, International Journal of Network Security, IJNS, Volume 14, No 1, Jan 2012.

13) Design Methodology of Solar Powered Load Controller using Sensor Assisted Remote Battery State of Charge Estimator, ASME International Mechanical Engineering Congress & Exposition, Vancouver, British Columbia, November 2010. DOI: 10.1115/IMECE2010-37008.

14) Sender-Side Public Key Deniable Encryption Scheme, International Conference on Advances in Recent Technologies in Communication and Computing, ARTCom, Kottayam, India, October 2009, IEEE Xplore DOI: 10.1109/ARTCom.2009.107.

15) Deniable Encryption in Replacement of Untappable Channel to Prevent Coercion, The Second International Conference on Network & Communications Security, Chennai, India, December 2010, DOI:10.1007/978-3-642-17878-8_50.

16) A New Approach on Congestion Control with Delay Reduction in Mobile Networks, International Conference on Advances in Recent Technologies in Communication and Computing, ARTCom, Kottayam, India, October 2010. IEEE Xplore DOI: 10.1109/ARTCom.2010.45.

17) A New Approach on Cluster based Call Scheduling for Mobile Networks, Journal of Advances in Information Technology, Vol 3, No 3 (2012), 184-190, Aug 2012, DOI: 10.4304/jait.3.3.184-190.

18) Uncoercibility in e-voting and e-auctioning mechanisms using deniable encryption, International Journal of Network Security and Its Applications, IJNSA, Vol 3, Number 2, March 2011.

19) A Symbolic Framework for Recognizing Activities in Full Motion Surveillance Videos, IEEE SSCI 2016.

20) Deploying a quantum annealing processor to detect tree cover in aerial imagery of California, PLOS One, 2017.

Patents

1) Computer Implemented System and Method for High performance Visual Tracking, US Provisional Patent Application No. 61/711,102, filed October 8, 2012.


Book Chapters

1) Deep Learning for Very High Resolution Imagery Classification, Taylor and Francis Machine Learning Book Chapter (under review).

2) Unmixing Algorithms: A Review of Techniques for Spectral Detection and Classification of Land Cover from Mixed Pixels on NASA Earth Exchange, Large Scale Machine Learning in the Earth Sciences, CRC Press, Taylor and Francis Group, Chapman & Hall (under review).


Recent Awards

1) AWS Climate Research Grant - Awarded a grant of ~1 Million core hours for performing climate research in support for the President's White House climate action plan, October 2014.

2) AWS Hackathon - Won 1st prize at AWS Hackathon, December 2014.


Public Datasets



N-MNIST Dataset


Noisy Bangla Dataset


SAT-4 and SAT-6 Datasets


MAPTrack - Demo Videos










Event Chaining - Demo Videos