Have you ever seen an Avocado-shaped armchair? (or something even more bizarre) Is it possible to generate such images? Have you heard of DALL·E, which creates images from text prompts?
This project aims to generate realistic 3D object renders (generating the geometry and color of a wide range of objects without 3D supervision) from textual prompts. Future works could include focusing on a particular domain like healthcare (MRI Scans) etc.
Learning outcomes: Generative Modeling, Neural Radiance Fields, Exploiting Large Pre-Trained Models - CLIP(Contrastive Language-Image Pre-Training)