Do you want to create a video just by entering a piece of text? That's the thing Sora, a new AI engine from OpenAI recently announced, can do it. Let's learn about Sora through this article!
| Join the channel Telegram belong to AnonyViet 👉 Link 👈 |
What is Sora?
Sora is a Diffusion model based on transformer architecture, developed by OpenAI to convert from text to video. Sora uses the annotation technique again introduced with Dall-E 3, based on a sample user prompt to create a highly detailed descriptive prompt. Not only creating videos from text, Sora can also create videos from still images, make them move and adjust the frame according to the video format.
What makes Sora outstanding?
Sora has many unique features not found in any other model. Here are some features:
- High quality videos (up to 1080p) can be created from just text snippets. It follows user prompts very well and depicts a dynamic physical world.
- Create AI videos up to one minute, much longer than current text-to-video models that can only create 3-4 second videos.
- Has a deep understanding of language and can create “compelling characters that express vivid emotions.” It can also create multiple different shots in one video with consistent characters and scenery throughout the video.
- Use Neural Radiation Field (NeRF) to create 3D scenes from 2D images.
What can Sora do?
Here are some examples of what Sora can do:
- Create an introductory video about a product, service, company, or individual by simply entering a text description about it.
- Create a video tutorial, lecture, or presentation on a topic simply by entering a text that outlines the steps, content, or idea to be conveyed.
- Create a narrative, poetic, or musical video simply by entering a text with the desired structure, rhythm, or music.
- Create a video that transforms, changes, or interacts with a still image simply by entering a text describing the desired action, effect, or result.
Where can Sora be experienced?
Sora is a model that has not been made public to everyone. OpenAI is working with scientists to test whether the model is harmful or dangerous. They also allowed a number of people in the film, design and art fields to use Sora to provide comments and perfect the model before its official launch.
Epilogue
Sora is a groundbreaking text-to-video AI tool from OpenAI, bringing a breath of fresh air to the AI industry. Sora can open up new creative possibilities and applications for users. Let's look forward to great videos created by Sora in the future.
