StoryTime creates imaginations into short stories. Need a quick bedtime story? Use your voice and describe your current mood or interests. StoryTime will creatively generate a short story. Create a new story or go back to your favorites anytime!
encode-ai-final-demo-480-short.mp4
In our project, we employed three AI APIs to generate stories, summaries, and accompanying images based on narrative prompts. Here's a breakdown of our approach:
We utilized the OPENAI Assistant API to generate engaging narratives. By assigning the assistant the persona of a renowned sci-fi writer, we ensured immersive storytelling experiences.
Leveraging the capabilities of the OPENAI Chat API, we extracted concise summaries from the generated sci-fi stories.
To complement our stories, we employed the DALLE-3 model to produce visual representations.
Initially, our images reflected a traditional sci-fi aesthetic, which we found to be outdated.
Transitioning to a cyberpunk theme introduced visually striking elements, though facial features remained inconsistent and sometimes distorted.
Despite efforts to enhance character details, challenges persisted in achieving satisfactory facial rendering.
Implementing an animated theme proved successful, as it offered greater tolerance for unconventional facial features.
Finally, we refined our images by removing extraneous text, ensuring clarity and focus on visual storytelling.
- NextJs / Typescript
- Tailwind / DaisyUI / React Aria Components
- OpenAI
From root directory
npm install
-
Rename
.env.exampleto.env.local -
Include your Open AI key for
OPENAI_API_KEY=in the env file- Sign up for an api key at OpenAI
-
(Optional) Include an
ASSISTANT_ID=in the env file. A new assistant will be created if one is not provided.
Run command in terminal of root directory:
npm run dev
Open link in browser (Default link): http://localhost:3000/
- 👨💻 Chris Salvador
- 👩💻 Vivien Zhou
- 👨💻 Ahmed Shalabi