This project aims to generate photorealistic object and terrain models for dynamic immersive environments and/or the creation and augmentation of datasets for image recognition. Objects and terrain could include urban areas, buildings, vehicles, aircraft, ships, and other objects.
This project involves identifying a valid approach and relevant dataset, constructing and training a model, and conducting experiments to assess accuracy and performance. Alternatively, an existing approach could be integrated into a game environment for dynamic terrain and object generation.
Potential approaches include generative deep learning, such as a 3D generative adversarial network (GAN): https://dl.acm.org/doi/10.1145/3373017.3373067
This project involves investigating a specific approach of your choosing related to generating or recognising photorealistic 3D models. If you would like to participate in this project, please send me a 1-2 paragraph proposal for your specific approach.