OpenAI announces the new text-to-video generation model called “Sora”. Sora is an AI model that will help you to create realistic videos from text prompts.
Sora can generate videos up to a minute long while maintaining the quality of the video prompt.
Sora can create complex situations with multiple moving characters, objects, and precise background and subject details.
The model can understand your prompt easily and understand the background of the prompt that you have asked to generate more accurate results like the real world.
The model also has a deep understanding of languages which enables the prompts to be more accurate and generate the videos you want.
The model is not perfect. It has many flaws which you can see generating more complex scenes.
How to get Access to Sora?
This model is not being released to the public to use. It is in the research stage to make things clear to release to the public.
They are working with red teams to find flaws in the model that can be hateful content, bias, sexually explicit content, hateful imagery, and other IPs.
How videos are generated?
To understand in simple language, Sora is a diffusion model that generates a video frame by frame with many steps. It started as noisy like when the TV isn’t tuned to a channel and with many steps it gradually removes the noise and creates a real video.
Sora can generate entire videos all at once or extend generated videos to make them longer.
10 interesting videos generated by Sora.
1/ A stylish woman walks down a Tokyo street
2/ Close-up shot of a Victoria-crowned pigeon
3/ Snowy Tokyo city
4/ Historical footage of California during the Gold Rush
5/ A movie trailer featuring the adventures of the 30-year-old spaceman wearing a red wool
6/ Extreme close-up of a 24-year-old woman’s eye blinking
7/ Drone view of waves crashing against the rugged cliffs along Big Sur’s Garay Point beach
8/An instructional cooking session for homemade gnocchi hosted by a grandmother social media influencer
9/ A wizard wearing a pointed hat and a blue robe with white stars casting a spell
— Soubhagya Sahu (@soubhagyaside) February 16, 2024
10/ Dogs podcasting in mountain
Read more:
