Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Page Not Found

Page not found. Your pixels are in another canvas. Read more

</article> </div>

Gradients - Ritwik Gupta

About me Read more

</article> </div>

Page not in menu

This is a page not in th emain menu Read more

</article> </div>

Posts

How to Write an Impactful Statement of Purpose

12 minute read

Published:

As the grad school application season unfolds, students worldwide, including yourself, are diligently preparing their applications. The statement of purpose (SOP) stands out as the most crucial and anxiety-inducing component of this process. Read more

</article> </div>

The Department of Defense is Prioritizing Open Source Software. Here’s How Open Source Projects Can Benefit.

10 minute read

Published:

On January 26, 2022, the new Chief Information Officer (CIO) of the U.S. Department of Defense (DoD), John B. Sherman, released a memo to the entire Department titled “Software Development and Open Source Software”. In this memo, the CIO addresses two primary concerns: 1) using open source software (OSS) introduces supply chain risks for DoD software programs, and 2) sharing DoD code via open source channels without proper checks enables potential leaks of proprietary DoD information to adversaries. In laying out how these two concerns should be addressed properly, the CIO categorizes OSS into a unique position, one which can be utilized by OSS foundations and project maintainers to gain funding for their essential contributions. Read more

</article> </div>

Thoughts About My High School’s Math Curriculum

10 minute read

Published:

I went to high school to Thomas Jefferson High School in Pittsburgh, PA (not the famous one in Alexandria, VA). I wrote an email to my math teachers there talking about my thoughts about our high school math curriculum now that I’ve been through an undergrad, currently taking grad courses, and have a full-time job doing math for a living. I’m pasting the emails as-is here, with some additional formatting. Please let me know if you have any comments or criticisms. Read more

</article> </div>

Exploring Deep Learning: Theory and Practice

1 minute read

Published:

This is a page for my talk given at the CMU Data Science Club at Doherty A302 on October 5th, 2017. Read more

</article> </div>

Making a Self Driving Car - Part 1

8 minute read

Published:

I just graduated from the University of Pittsburgh and have some time to relax before starting my job, so my dad and I decided to build a self-driving car. We think we can do it, and even if we can’t, at least we’ll learn some cool stuff while doing it. The goal is to start simple and build as we go. Read more

</article> </div>

Understanding HL7 Structure

6 minute read

Published:

What is HL7

HL7, more formally known as Health Level-7, is a standard used in the healthcare industry to transfer clinical and administrative data. Much like many SaaS services “speak JSON”, healthcare applications “speak HL7”. Read more

</article> </div>

Idea for Efficient Cell Image Compression

1 minute read

Published:

Premise

Pathology and histology slide images are taken with extremely high resolution cameras, resulting in:

  1. High cost of storage
  2. Bandwidth issues while transporting images

It is important to note that these images often rely on lossless compression, because any artifacting will result in a lowered ability for the doctor to give accurate results.

Example of a cell image.

Proposed Solution

In most regions of the body, neigboring cells looks alike to the cells bordering it. I propose the following solution to compress the images:

  1. Segment the image into distinct cells.
  2. Take a dot product of the matrix of cells with itself (ATA) to figure out which cells are most similar to other cells. I call these “reference cells”.
  3. Store a set of deltas for all other cells in terms of rotations, translation, and transformations.
  4. Apply further compression using any standard algorithm to the deltas themselves.

Example of “reference” cells.

A set of deltas in a binary format would be better than a large amount of pixels and allow for better compression.

I attempted to do something along these lines on GitHub here, but had to stop development due to time constraints. Hopefully I can continue this down the line! Read more

</article> </div>

Making an API for Pitt

8 minute read

Published:

Making an API for Pitt Read more

</article> </div>

portfolio

Portfolio item number 1

Published:

Short description of portfolio item number 1
Read more

</article> </div>

Portfolio item number 2

Published:

Short description of portfolio item number 2
Read more

</article> </div>

publications

Open Problems in Robotic Anomaly Detection

Ritwik Gupta, Zachary T. Kurtz, Sebastian Scherer, Jonathon M. Smereka

Venue: arXiv preprint

Motivated by the development of ROS 2, this work discusses open problems in the field of robotic anomaly detection and presents an inverse reinforcement learning-based approach to detecting anomalous motion.

Focusing on the Big Picture: Insights into a Systems Approach to Deep Learning for Satellite Imagery

Ritwik Gupta, Carson D. Sestili, Javier A. Vazquez-Trejo, Matthew E. Gaston

Venue: 2018 IEEE International Conference on Big Data

Focused around the IARPA Functional Map of the World Challenge, this work discusses how to scale deep learning at an academic lab for geospatial analysis.

Creating xBD: A Dataset for Assessing Building Damage from Satellite Imagery

Ritwik Gupta, Bryce Goodman, Nirav Patel, Ricky Hosfelt, Sandra Sajeev, Eric Heim, Jigar Doshi, Keane Lucas, Howie Choset, Matthew Gaston

Venue: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019

Preliminary work discussing xBD, the foundational dataset for assessing building damage after natural disasters from very-high resolution satellite imagery with over 850,000 annotations across 45,000 square kilometers.

xBD: A Dataset for Assessing Building Damage from Satellite Imagery

Ritwik Gupta, Richard Hosfelt, Sandra Sajeev, Nirav Patel, Bryce Goodman, Jigar Doshi, Eric Heim, Howie Choset, Matthew Gaston

Venue: arXiv preprint

The foundational dataset for assessing building damage after natural disasters from very-high resolution satellite imagery with over 850,000 annotations across 45,000 square kilometers.

Region-level Active Detector Learning

Michael Laielli, Giscard Biamby, Dian Chen, Ritwik Gupta, Adam Loeffler, Phat Dat Nguyen, Ross Luo, Trevor Darrell, Sayna Ebrahimi

Venue: arXiv preprint

A new strategy that subsumes previous Image-level and Object-level approaches into a generalized, Region-level approach.

WiSoSuper: Benchmarking Super-Resolution Methods on Wind and Solar Data

Rupa Kurinchi-Vendhan, Björn Lütjens, Ritwik Gupta, Lucien Werner, Dava Newman

Venue: Climate Change AI Workshop at NeurIPS 2021

An extensible benchmark of deep learning-based super-resolution techniques on wind and solar data. We accompany the benchmark with a novel public, processed, and machine learning-ready dataset for benchmarking super-resolution methods on wind and solar data.

Snowpack Estimation in Key Mountainous Water Basins from Openly-Available, Multimodal Data Sources

Malachy Moran, Kayla Woputz, Derrick Hee, Manuela Girotto, Paolo D'Odorico, Ritwik Gupta, Daniel Feldman, Puya Vahabi, Alberto Todeschini, Colorado J Reed

Venue: CVPR 2022 Workshop on Multimodal Learning for Earth and Environment

We fuse satellite and weather data to estimate snowpack depth in key mountainous regions and beat single-source estimation by 5.0 inches RMSE.

Satlas: A Large-Scale, Multi-Task Dataset for Remote Sensing Image Understanding

Favyen Bastani, Piper Wolters, Ritwik Gupta, Joe Ferdinando, Aniruddha Kembhavi

Venue: International Conference on Computer Vision (ICCV) 2023

A foundational remote sensing dataset with over 290M labels under 137 categories and seven label modalities for pre-training large machine learning models.

xView3-SAR: Detecting Dark Fishing Activity Using Synthetic Aperture Radar Imagery

Ritwik Gupta*, Fernando Paolo*, Tsu-ting Tim Lin*, Bryce Goodman, Nirav Patel, Daniel Kuster, David Kroodsma, Jared Dunnmon

Venue: NeurIPS 2022

The largest labeled dataset for training ML models to detect and characterize vessels and ocean structures in synthetic aperture radar imagery. xView3 is built to help control illegal, unreported, and unregulated fishing.

Emerging Technology and Policy Co-Design Considerations for the Safe and Transparent Use of Small Unmanned Aerial Systems

Ritwik Gupta, Alexander Bayen, Sarah Rohrschneider, Adrienne Fulk, Andrew Reddie, Sanjit A. Seshia, Shankar Sastry, Janet Napolitano

Venue: Center for Security in Politics, UC Berkeley

With the meteoric rise of small unmanned aerial systems, we discuss policy shortcomings in integrating sUAS technology in a safe fashion into our society. We suggest technology and policy co-design approaches to addressing these gaps in our systems.

Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learning

Ritwik Gupta*, Colorado Reed*, Shufan Li*, Sarah Brockman, Christopher Funk, Brian Clipp, Kurt Keutzer, Salvatore Candido, Matt Uyttendaele, Trevor Darrell

Venue: International Conference on Computer Vision (ICCV) 2023

A pre-training method to make encoders robust to imagery captured at varying satellite resolutions. State-of-the-art multi-scale pre-training method and the largest satellite imagery foundation model, to date.

Orbital hypersonic delivery systems threaten strategic stability

Ritwik Gupta

Venue: The Bulletin of Atomic Scientists

We assess that China's development of a fractional orbital hypersonic delivery system, combining hypersonic glide vehicles with orbital bombardment, presents a concerning challenge to global stability, allowing for faster, undetectable delivery of large nuclear payloads and signaling renewed interest in first-strike capabilities.

ClimSim: An open large-scale dataset for training high-resolution physics emulators in hybrid multi-scale climate simulators

Sungduk Yu, Walter M. Hannah, Liran Peng, Mohamed Aziz Bhouri, Ritwik Gupta, Jerry Lin, Björn Lütjens, Justus C. Will, Tom Beucler, Bryce E. Harrop, Benjamin R. Hillman, Andrea M. Jenney, Savannah L. Ferretti, Nana Liu, Anima Anandkumar, Noah D. Brenowitz, Veronika Eyring, Pierre Gentine, Stephan Mandt, Jaideep Pathak, Carl Vondrick, Rose Yu, Laure Zanna, Ryan P. Abernathey, Fiaz Ahmed, David C. Bader, Pierre Baldi, Elizabeth A. Barnes, Gunnar Behrens, Christopher S. Bretherton, Julius J. M. Busecke, Peter M. Caldwell, Wayne Chuang, Yilun Han, Yu Huang, Fernando Iglesias-Suarez, Sanket Jantre, Karthik Kashinath, Marat Khairoutdinov, Thorsten Kurth, Nicholas J. Lutsko, Po-Lun Ma, Griffin Mooers, J. David Neelin, David A. Randall, Sara Shamekh, Akshay Subramaniam, Mark A. Taylor, Nathan M. Urban, Janni Yuval, Guang J. Zhang, Tian Zheng, Michael S. Pritchard

Venue: Neural Information Processing Systems (NeurIPS) 2023

The largest-ever dataset designed for hybrid ML-physics research. It comprises multi-scale climate simulations, developed by a consortium of climate scientists and ML researchers.

Accelerating the Evolution of AI Export Controls

Ritwik Gupta and Andrew Reddie

Venue: Tech Policy Press

Current US AI hardware export controls are based on the best AI accelerator chip available at that time. This presents wide loopholes which allow adversarial nations to still maintain their capabilities. We propose an alternate way to set export control thresholds based on the analysis of specific ML workloads.

Proliferate, Don’t Obliterate: How Responsive Launch Marginalizes Anti-Satellite Capabilities

Ritwik Gupta and Andrew Reddie

Venue: War on the Rocks

We analyze how the emerging responsive launch industry fundamentally shifts the strategic calculus of ASAT weapons.

See, Say, and Segment: Teaching LMMs to Overcome False Premises

Tsung-Han Wu, Giscard Biamby, David Chan, Lisa Dunlap, Ritwik Gupta, Xudong Wang, Joseph E. Gonzalez, Trevor Darrell

Venue: In review

A method to prevent large, multimodal models (LMMs) to stop hallucinating when given false premises.

LAION and the Challenges of Preventing AI-Generated CSAM

Ritwik Gupta

Venue: Tech Policy Press

I examined the challenges in preventing AI-generated Child Sexual Abuse Material (CSAM), such as within the widely-used LAION-5B dataset, emphasizing the need for updated legal and technological strategies to tackle the spread of such content by generative AI technologies.

talks

Talk 1 on Relevant Topic in Your Field

Published:

This is a description of your talk, which is a markdown files that can be all markdown-ified like any other post. Yay markdown! Read more

</article> </div>

Tutorial 1 on Relevant Topic in Your Field

Published:

More information here Read more

</article> </div>

Talk 2 on Relevant Topic in Your Field

Published:

More information here Read more

</article> </div>

Conference Proceeding talk 3 on Relevant Topic in Your Field

Published:

This is a description of your conference proceedings talk, note the different field in type. You can put anything in this field. Read more

</article> </div>

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post. Read more

</article> </div>

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post. Read more

</article> </div>