imGAN
OVERVIEW

Built a Machine learning pipeline for training, classifying, and detecting synthetic images generated by GANs. Focused on identifying visual artifacts to distinguish real images from AI-generated content, using convolutional networks and custom classifier models.

faces generated

YEAR

2021

DOMAIN

Generative AI
Machine Learning

TECH STACK

Python

PyTorch

Jupyter

About the project

Explored generative modeling as an unsupervised learning task to synthesize new, realistic data samples. Leveraged GANs to capture underlying data distributions and reproduce images indistinguishable from the original dataset, showcasing model generalization and pattern learning.

Smooth Scroll
This will hide itself!
imGAN
OVERVIEW

Built a Machine learning pipeline for training, classifying, and detecting synthetic images generated by GANs. Focused on identifying visual artifacts to distinguish real images from AI-generated content, using convolutional networks and custom classifier models.

faces generated

YEAR

2021

DOMAIN

Generative AI
Machine Learning

TECH STACK

Python

PyTorch

Jupyter

About the project

Explored generative modeling as an unsupervised learning task to synthesize new, realistic data samples. Leveraged GANs to capture underlying data distributions and reproduce images indistinguishable from the original dataset, showcasing model generalization and pattern learning.

Smooth Scroll
This will hide itself!
imGAN
OVERVIEW

Built a Machine learning pipeline for training, classifying, and detecting synthetic images generated by GANs. Focused on identifying visual artifacts to distinguish real images from AI-generated content, using convolutional networks and custom classifier models.

faces generated

YEAR

2021

DOMAIN

Generative AI
Machine Learning

TECH STACK

Python

PyTorch

Jupyter

About the project

Explored generative modeling as an unsupervised learning task to synthesize new, realistic data samples. Leveraged GANs to capture underlying data distributions and reproduce images indistinguishable from the original dataset, showcasing model generalization and pattern learning.

Smooth Scroll
This will hide itself!