As part of OCI Generative AI Research Group, our objective is to create a predictive model that can estimate render times for a given task on a specific compute shape with an accuracy of <15%. Achieving this will add efficiency to render planning, tracking, and monitoring workflows in the Animation Media Production process. This work complements our broader R&D charter, which includes:

• Research: Conducting in-depth research on image and video generation techniques, including generative adversarial networks (GANs), diffusion models, and other emerging approaches.

• Model Development: Designing, developing, and training state-of-the-art generation models that meet the highest quality standards.

• Team Leadership: Building and mentoring a high-performing team of scientists and engineers to execute on ambitious technical goals.

• Collaboration: Partnering with cross-functional teams to integrate advanced generation capabilities into applications and products.

• Innovation: Identifying new opportunities for generation technology and exploring emerging trends in the space.

• Industry Awareness: Staying current with advancements in both image and video generation, ensuring our approaches remain cutting-edge.