RadiusAI

Photo-Realistic Synthetic Image Generation

Radius AI is on the forefront of AI-driven retail and healthcare computer vision-driven systems. As part of our retail expansion at the time of this project, Radius AI was categorizing an ever-expanding list of consumer products. For this project, this student team set out to investigate, research, and develop state-of-the art methods for photo-realistic synthetic image generation. This student team also worked to propose solutions ranging from AI/GAN based approaches to 3D reconstruction, implement object detection and recognition, and demonstrate any resulting improvements. Desired outcomes and deliverables this student team worked towards included creating a proven methodology that may be used to scale-ably create diverse training scenes. This student team worked to: a) Deliver a report or presentation discussing all methodologies that were attempted b) Deliver the resulting codebase in the form of a github that may be utilized by RadiusAI