Compilation of Text to Video models

several approaches are available for generating videos from the given text prompt. The existing methodologies encompass Generative Adversarial Networks (GANs), Transformer architectures, Variational Auto encoders (VAEs), Stable Diffusion techniques, and 3D Convolutional Neural Networks (3D CNNs) are not up to the mark. In this research article methodologies explored are Sora, Tune-a-video, Cog Video, and ModelScopeT2V. Out of these ModelScopeT2V performs in a better way leverages stable diffusion and transformers to generate semantically coherent videos directly from text . It processes input text, extracts features, refines latent space representation, and converts it into video frames. Spatiotemporal blocks and attention mechanisms ensure both visual appeal and temporal consistency in the generated videos.