Arjun Mani

profile pic

Hi there! 👋 I am a PhD student in Computer Science at Columbia University, where I work in machine learning. I am broadly interested in developing AI systems that can understand, predict, and design the physical world. My research interests include learned physical simulation, computational design, 3D learning and representations, and AI4Science. I'm fortunate to be advised by Carl Vondrick and Rich Zemel, and supported by an NSF Fellowship.

Previously, I completed my Bachelors in Computer Science at Princeton University, with a minor in Applied Math. I had the pleasure of working with Olga Russakovsky in the Princeton Visual AI Lab and Ryan Adams. I also spent a summer at Google.

I am trained in Indian Classical Music and perform concerts. Music page coming soon!

linkedin github email email

Research

SurfsUp: Learning Fluid Simulation for Novel Surfaces

Arjun Mani*, Ishaan Preetam Chandratreya*, Elliot Creager, Carl Vondrick, Richard Zemel

ICCV 2023.

[website] / [arXiv] / [code]


We propose a learned model for fluid-surface simulation that models solid surfaces implicitly, allowing us to generalize considerably to novel objects and scenes. We also demonstrate more efficient simulation, and ability to solve inverse design problems.

Point and Ask: Incorporating Pointing into Visual Question Answering

Arjun Mani, Will Hinthorn, Nobline Yoo, Olga Russakovsky

VQA Workshop, CVPR 2021 (Poster Spotlight).

[arXiv] / [code]


We extend Visual Question Answering (VQA) to questions requiring pointing gestures and introduce benchmark datasets and model designs for this new question space.

Representing Words in a Geometric Algebra

Arjun Mani, Ryan Adams

Best Overall Project, Princeton Program in Applied Mathematics (PACM)

[Report] / [Press release]


Represented word embeddings using multivectors in a geometric algebra and showed results suggesting greater expressivity in word similarity, analogy solving.

Adaptive Estimation of Active Contour Parameters Using Convolutional Neural Networks and Texture Analysis

IEEE Transactions on Medical Imaging, vol. 36, no. 3, March 2017

Assaf Hoogi, Arjun Subramaniam*, Rishi Veerapaneni*, Daniel Rubin

[Paper] / [IEEE page]


Pioneered a deep learning-based, adaptive improvement of state-of-the-art segmentation techniques, which achieved ~10% accuracy gains on MRI and CT lesion datasets.

Other

Check out my Projects page for various projects that I have worked or am working on, and my music page for links to past and upcoming performances. I will be starting up a blog soon as well, so keep an eye out for that!

Lastly, this website template owes thanks to Sharon Zhang.