Using SORA extends beyond just text inputs; you can also generate videos by submitting image inputs.
When you send an image input to SORA along with a prompt detailing your request, SORA can animate or transform the image into a video based on the instructions provided.
Examples of Generating Videos from Images with SORA
Below are the examples when submitting image inputs to OpenAI’s SORA AI model, accompanied by prompts:
#1 Image Input
Prompt: A Shiba Inu dog wearing a beret and black turtleneck.
#2 Image Input
Prompt: Monster Illustration in flat design style of a diverse family of monsters. The group includes a furry brown monster, a sleek black monster with antennas, a spotted green monster, and a tiny polka-dotted monster, all interacting in a playful environment.
#3 Image Input
Prompt: An image of a realistic cloud that spells “SORA”.
#4 Image Input
Prompt: In an ornate, historical hall, a massive tidal wave peaks and begins to crash. Two surfers, seizing the moment, skillfully navigate the face of the wave.
You can learn more on OpenAI research paper.
How to Use SORA with Image Inputs?
While detailed usage instructions, including client interfaces and API access for SORA, have not been released, the process is anticipated to be similar to using OpenAI’s other models like ChatGPT and DALL·E.
For API usage, you would simply add the image input as a parameter when making your API call.
For client-based interactions, it can be compared to uploading a file in ChatGPT: you upload your image, submit your prompt, and receive a video generated from these inputs.
Also Read: How to Use Sora With Video Input
Limitations of Using SORA with Image Inputs
When utilizing SORA with image inputs, the existing limitations of the SORA remain applicable. Here are the key limitations to keep in mind:
- Video Length Limit: SORA can generate videos with a maximum length of one minute. This limitation applies regardless of whether the input is text, image, or video, affecting the duration of content you can create.
- Video Quality Cap: Videos produced using SORA, including those generated from image inputs, are capped at 1080p resolution. This ensures a high-definition output but does not allow for higher resolutions that might be desired for certain applications.
- Content Restrictions: As with all OpenAI models, SORA adheres to strict content guidelines to ensure ethical usage. Therefore, it is not possible to generate videos containing sexual content, excessive violence, or celebrity likenesses using SORA. These safety measures are in place to prevent misuse and protect intellectual property rights.
Despite these limitations, using SORA with image inputs opens up a realm of possibilities for creative video production.
Can SORA Generate Images?
Yes, SORA is also capable of generating images, much like OpenAI’s DALL·E 3 model. This means SORA isn’t limited to producing videos; it can also create images.
The resolution of images produced by SORA can vary, with capabilities up to 2048×2048 pixels, offering high-quality visual content.
Prompt: A snowy mountain village with cozy cabins and a northern lights display, high detail and photorealistic dslr, 50mm
This versatility makes SORA a powerful tool not only for video creation but also for generating static images, broadening the scope of content that can be produced with this advanced AI model.