Vinay Uday Prabhu's Projects
Testing
Generating adversarial examples using style transfer
Documenting the nature & depth of art-websites scraped in CommonCrawl
This is for my contribution to the BYO Art Community Gallery Night at the Make Factory, SF
A home for all my neural art horrors
Image augmentation library in Python for machine learning.
In this work, I aspire to harness the fast, arbitrary artistic neural style transfer technique of [1], to sample images from Raja Ravi Varma’s body of work and transfer it to the avant garde Bengal school of art style captured in Abanindranath Tagore’s paintings. I used the open-sourced implementation of [1] that is part of the ’Magenta’ project available at [2]. References: [1] https://arxiv.org/pdf/1705.06830.pdf [2] https://github.com/tensorflow/magenta/tree/master/magenta/models/arbitrary_image_stylization
Beyond the Imitation Game collaborative benchmark for enormous language models
My burning man code and data
My code for chip and terra for burning-man 2019
Multi-class dataset splitting so as to maintain the class-wise ratios
Here reside my colab scratches
A survey of how model confidence is exposed to ML developers
⛵️ Implementation a variety of popular Image Classification Models using TensorFlow2. [ResNet, GoogLeNet, VGG, Inception-v3, Inception-v4, MobileNet, MobileNet-v2, ShuffleNet, ShuffleNet-v2, etc...]
JFC! What a hot mess. *Scream into void*
Reproducible open-sourced digital art.
Estimate the total emissions for popular CryptoArt platforms.
All of the ImageNet meta-datasets and notebooks
Ushering in deep learning's foray into the netherworld of eyeballing
Deterministic initialization of Deep Neural Networks
On disentangling the menagerie of disentanglement papers
The world's simplest facial recognition api for Python and the command line
Collection of my favorite Python packages from 2020
Here's where all my Colab Pro abuse sits. I foresee zero worth in these
An exploration into transfer learning and adversarial perturbations
Code for 'Vulnerability of deep learning based gait biometric recognition to adversarial perturbations'
Accelerometric time series transfer learning
How do ML Lang_ID classifiers handle human generated 'artistic' gibberish