Effortless Image Captioning with TensorFlow.js

Generate image captions directly in your web or Node.js projects using pre-trained TensorFlow.js models. Ideal for web developers, designers, and researchers to add visual understanding and accessibility to their applications.

Key Features:

Simple API: Get captions with one line of code.
Pre-trained Model Flexibility: Choose from different pre-trained models for varying accuracy and performance needs.
Multiple Image Formats: Supports common image formats like JPEG, PNG, and more.
Code Examples: Learn how to integrate the package into your projects with clear examples.

Installation:

npm install tfjs-image-captioner

Usage

import imageCaptioning from 'tfjs-image-captioner';

const imagePath = 'path/to/your/image.jpg';
imageCaptioning.getImageCaptions(imagePath)
  .then(captions => {
    console.log('Captions:', captions);
  })
  .catch(error => {
    console.error('Error:', error);
  });

For more details, examples, and advanced usage, please refer to the documentation:

Documentation: https://www.tensorflow.org/js/models
Community Forum: https://stackoverflow.com/questions/tagged/tensorflow.js

Try different pre-trained models:

For future version consideration integrate different pre-trained models with varying accuracy and performance characteristics.

Model Selection API
Adaptive Model Loading

License:

MIT License

sdkdeepa / tfjs-image-captioner