Keshigeyan Chandrasegaran

Official Code for GPIC: A Giant Permissive Image Corpus for Visual Generation

GPIC is a Stanford-created 100-million-image training dataset with permissive licensing and AI-generated captions, plus an evaluation toolkit and baseline model for benchmarking image generation systems.

[NeurIPS 2024] Official code for HourVideo: 1-Hour Video Language Understanding

[NeurIPS 2025 Oral] Official Code for Exploring Diffusion Transformer Designs via Grafting

End-to-end Learning of Deep Visual Representations for Image Retrieval.

PASCAL VOC Object Classification: The goal of this project is to recognize objects from a number of visual object classes in realistic scenes.

A two stage multi-modal loss model along with rigid body transformations to regress 3D bounding boxes

No description.

A. N. A is a dynamic online gathering place where students can ask, answer and explore 24/7, with the guidance of instructors to promote a healthy collaborative dynamic learning experience powered by android platform.

Ethcert.io leverages on Ethereum’s framework to provide a smart contract certification service.

A set of useful components for common UI elements.

Code for ICML 2022 paper "Out-of-distribution Detection with Deep Nearest Neighbors"

The author's officially unofficial PyTorch BigGAN implementation.

Repo for our CVPR Paper: Watch your Up-Convolution: CNN Based Generative Deep Neural Networks areFailing to Reproduce Spectral Distributions

Library of deep learning models and datasets designed to make deep learning more accessible and accelerate ML research.

2-SAT is a classic computational problem of assigning Boolean values (true or false) to satisfy a system of constraints operating on pairs of variables. Though Boolean satisfiability problems are generally NP-complete, 2-SAT problems are a special instance of Boolean satisfiability problems that can be solved in linear time. I have used Kosaraju’s algorithm to solve the 2-SAT problem where the implementation has a linear time complexity and linear space complexity.

Boilerplate application for a Python/Flask JWT Backend and a Javascript/React/Redux Front-End with Material UI.

No description.

No description.

No description.

This is a pytorch re-implementation of Learning a Discriminative Filter Bank Within a CNN for Fine-Grained Recognition

No description.

Reveal CSS animation as you scroll down a page

No description.

No description.

8-bit ALU implemented on Mojo FPGA

Mojo implementation for restricted rock paper scissors game

No description.

No description.

No description.

No description.