I am a research scientist at SRI International’s Center for Vision Technologies in Princeton, NJ. I work on a wide range of machine learning applications centered around the application of deep learning to vision and language problems.

During my PhD I was advised by Dhruv Batra and worked closely with Devi Parikh at Georgia Tech. During my PhD I explored these interests by building and understanding neural networks as applied to vision and language.

I want to contribute to technically and philosophically challenging long term research projects. My perspective and skills would best contribute to a holistic approach to building or understanding AI that treats artifacts and understanding as primary goals and publications as secondary. If that sounds like the candidate you’re looking for then send me an email at michael dot a dot cogswell at gmail dot com and find my CV here.


Dialog without Dialog: Learning Image-Discriminative Dialog Policies from Single-Shot Question Answering Data
Michael Cogswell, Jiasen Lu, Devi Parikh, Stefan Lee, Dhruv Batra
(Forthcoming Work)

Emergence of Compositional Language with Deep Generational Transmission
Michael Cogswell, Jiasen Lu, Stefan Lee, Devi Parikh, Dhruv Batra
(arXiv 2019)

Diverse Beam Search: Decoding Diverse Solutions from Neural Sequence Models
Ashwin K Vijayakumar, Michael Cogswell, Ramprasath R. Selvaraju, Qing Sun, Stefan Lee, David Crandall, Dhruv Batra
(AAAI 2018)
[arXiv] [demo]

Grad-CAM: Why did you say that? Visual Explanations from Deep Networks via Gradient-based Localization
Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, Dhruv Batra
(ICCV 2017)
[arXiv] [code] [demo] [video]

Stochastic Multiple Choice Learning for Training Diverse Deep Ensembles
Stefan Lee, Senthil Purushwalkam, Michael Cogswell, Viresh Ranjan, David Crandall, Dhruv Batra
(NIPS 2016)

Reducing Overfitting in Deep Networks by Decorrelating Representations
Michael Cogswell, Faruk Ahmed, Ross Girshick, Larry Zitnick, Dhruv Batra
(ICLR 2016)

Why M heads are better than one: Training a diverse ensemble of deep networks (similar to Lee et. al. 2016)
Stefan Lee, Senthil Purushwalkam, Michael Cogswell, David Crandall, Dhruv Batra
(arXiv preprint)

Combining the best of graphical models and convnets for semantic segmentation
Michael Cogswell, Xiao Lin, Senthil Purushwalkam, Dhruv Batra
(arXiv preprint)