A Systematic Review of Generative AI for Text-to-Video Generation (T2V)

This project is a systematic literature review of recent research in generative AI for text-to-video (T2V) generation, with the goal of providing an up-to-date synthesis of the technical landscape, emerging trends, and open challenges in the field. Text-to-video generation has advanced rapidly due to improvements in diffusion-based models, transformer-style architectures, multimodal representation learning, and scaling of training data and compute. At the same time, the area is characterized by fast iteration cycles, heterogeneous evaluation practices, and uneven reporting (e.g., limited disclosure of training data, compute budgets, or safety mitigations), which makes it difficult for researchers and practitioners to compare methods and identify robust research directions. This review aims to consolidate the evidence and provide a structured, reproducible account of what is known and what remains unresolved.