• Worked with PlayStation AI R&D team on generating 3D Generation and Reconstructions using deep learning methods such as GANs, VAE and Implicit functions.

​

• Implemented a novel architecture for generating 3D mesh conditioned on text input.

​

• Extended the SDF based method with novel contrastive learning loss to achieve transformation and generation of chairs based on text tags. 

​

• Achieved SOTA qualitative results.