Skip to content

Latest commit

 

History

History
51 lines (35 loc) · 2.76 KB

README.md

File metadata and controls

51 lines (35 loc) · 2.76 KB
Raw

Unleash your Imagination: Craft compelling narratives from your Images using the magic of Gemini 1.5 Flash.

Image-to-text.mp4

Project Explanation:

This project harnesses the power of Gemini 1.5 Flash, Google's advanced large language model, to analyze images and generate insightful information based on user prompts. It works by:

  • Image Input: The user provides an image, which the project processes to extract relevant features and data.
  • Prompt Input: The user also supplies a prompt, guiding the type of information they want to generate from the image (e.g., "Describe the scene," "Identify objects," or "Write a creative story").
  • Gemini 1.5 Flash Integration: The processed image data and prompt are fed into Gemini 1.5 Flash, which uses its powerful understanding of language and context to produce the desired information.
  • Output: The generated information is presented to the user, offering insights, descriptions, or creative content based on the image and prompt.

Potential Applications:

  • Image Captioning: Automatically generate descriptive captions for images.
  • Object Recognition: Identify and label objects within images.
  • Scene Understanding: Provide detailed descriptions of the scene depicted in an image.
  • Creative Writing: Generate stories, poems, or other creative content inspired by images.
  • Accessibility: Help visually impaired users understand the content of images.

click here to view the Application.

How it Works

1. Environment Setup:

  • The GOOGLE-API-KEY environment variable needs to be set with your valid Gemini API key.
  • Necessary libraries are imported: streamlit, google.generativeai, os, textwrap, PIL (Pillow), and pathlib.

2. Streamlit App Initialization:

  • Sets the page title to "Gemini".
  • Creates a header with the title "Gemini Image to Text Application 🖼️ 🖥️ 📄".

3. User Input:

  • A text input field is provided for the user to enter a prompt (optional).
  • A file uploader allows the user to upload an image.
  • If an image is uploaded, it is displayed in the app.

4. Gemini Interaction:

  • The get_gemini_response function handles the communication with the Gemini 1.5 Flash model.
  • It takes the user's prompt and the uploaded image as input.
  • If a prompt is provided, it is included in the request to Gemini along with the image otherwise, only the image is sent to Gemini.
  • The generated text response from Gemini is returned.

5. Output Presentation:

When the "Do the Magic" button is clicked:

  • The get_gemini_response function is called to generate the text.
  • The generated text is displayed under the subheader "The Response is".