試験
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 1 / 65
I want to generate high-quality advertising footage of a product from text prompts; when continuity between frames is important, which approach is most suitable?
①Generate frame descriptions directly with a large language model.
②Variational autoencoder that generates single frames.
③Using traditional editing software in combination with templates.
④Learning optimal policies via reinforcement learning
⑤A diffusion-based generative model specialized for video generation.