My Projects

Semantic layout to photorealistic image

DESCRIPTION

The objective of this project was that given a semantic layout of a scene, we want to generate a photorealistic image of a scene that conforms to the input layout. The generated image was realistic in appearance compared to the real-world scenes.

Example

Input-Output Example

TECHNOLOGIES USED

Python, Pytorch, CUDA, csh scripting

TOOLS USED

Google Colab, Google Cloud Platform, Amazon Web Services, Cityscapes dataset & NYU indoor scene dataset

PROJECT PRESENTATION

Click to view PROJECT PRESENTATION

PROJECT REPORT

Click to view PROJECT REPORT

Previous Post in ABOUT-ME

Discussion & Comments