Exif AI

2024-10-01.00.00.31.mov

Read this in other languages: 简体中文

About

Exif AI is a powerful CLI tool designed to write AI-generated image descriptions and/or tags directly into the metadata of image files. This tool leverages advanced AI models to analyze image content and generate descriptive metadata, enhancing the accessibility and searchability of your images.

Usage Example

CLI

Without Installation

You can run Exif AI directly using npx without installing it globally:

npx exif-ai -i example.jpeg -a ollama

With Installation

If you have installed Exif AI globally, you can run it directly from the command line:

exif-ai -i example.jpeg -a ollama

Options

Required options:

• -a, --api-provider <value>: Name of the AI provider to use (ollama for Ollama, zhipu for ZhipuAI, google for Google Gemini, coze_bot for Coze Bot, openai for OpenAI).

Optional options:

• -T, --tasks <tasks...>: List of tasks to perform ('description', 'tag', 'face').
• -i, --input <file> Path to the input image file.
• -p, --description-prompt <text>: Custom prompt for the AI provider to generate description. Defaults to a generic image description prompt.
• --tag-prompt <text>: Custom prompt for the AI provider to generate tags. Defaults to a generic image tagging prompt.
• -m, --model <name>: Specify the AI model to use, if supported by the provider.
• -t, --description-tags <tags...>: List of EXIF tags to write the description to. Defaults to common description tags.
• --tag-tags <tags...>: List of EXIF tags to write the tags to. Defaults to common tags.
• -v, --verbose: Enable verbose output for debugging.
• -d, --dry-run: Preview AI-generated content without writing to the image file.
• --exif-tool-write-args <args...>: Additional ExifTool arguments for writing metadata.
• --provider-args <args...>: Additional arguments for the AI provider.
• -w, --watch <path>: Watch directory for new files to process.
• --avoid-overwrite: Avoid overwriting if EXIF tags already exist in the file.
• --ext <extensions...>: File extensions to watch. Only files with this extensions will be processed.
• --concurrency <number>: The numbers of files to process concurrently in watch mode.
• --face-group-ids <group...>: List of face group IDs to use for face recognition.
• --repeat: The number of times to repeat the task if the AI-generated result is deemed unacceptable. This parameter helps ensure the quality of the output by allowing multiple attempts. Default value is 0. An AI-generated description is considered acceptable if it has more than 10 characters and is not in markdown format. AI-generated tags are considered acceptable if there are more than 1 tag and they are not in markdown format. Using this parameter will consume more tokens, which may incur additional costs. Use it at your own risk.

Example usage:

exif-ai -i example.jpg -a ollama -p "Describe this landscape photo."

Library

To use Exif AI as a library in your project, import it and use the provided functions:

import { execute } from "exif-ai";

const options = {
  tasks: ["description"], // List of tasks to perform
  path: "example.jpg", // Path to the input image file
  provider: "ollama", // Name of the AI provider to use
  descriptionTags: [
    "XPComment",
    "Description",
    "ImageDescription",
    "Caption-Abstract",
  ], // List of EXIF tags to write the description to
  tagTags: ["Subject", "TagsList", "Keywords"], // List EXIF tags to write the tags to
  descriptionPrompt: "Describe this landscape photo.", // Custom prompt for the AI provider to generate description
  tagPrompt: "Tag this image based on subject, object, event, place.", // Custom prompt for the AI provider to generate tags
  verbose: false, // Enable verbose output for debugging
  dry: false, // Preview AI-generated content without writing to the image file
  writeArgs: [], // Additional ExifTool arguments for writing metadata
  providerArgs: [], // Additional arguments for the AI provider
  avoidOverwrite: false, // Avoid overwriting if EXIF tags already exist in the file
  doNotEndExifTool: false, // Do not end ExifTool process after writing metadata
  faceGroupIds: [], // List of face group IDs to use for face recognition
  repeat: 0, // The number of times to repeat the task if the AI-generated result is deemed unacceptable
};

execute(options)
  .then(() => {
    console.log("Image description has been written to EXIF metadata.");
  })
  .catch((error) => {
    console.error("An error occurred:", error);
  });

Installation

To install Exif AI globally, use the following command:

npm install -g exif-ai

Tasks

Description

The description task generates a description of the image using the AI provider. The description is written to the specified EXIF tags defined in descriptionTags.

Tag

The tag task generates tags for the image using the AI provider. The tags are written to the specified EXIF tags defined in tagTags.

Face Recognition

The face task performs face recognition on the image using the Tencent Cloud API. The face recognition results are written to the specified EXIF tags defined in tagTags.

Currently, the face task requires user to enable face recognition service on Tencent Cloud and set a pair of Tencent Cloud API Secret ID and Tencent CLoud API Secret Key in the environment variable.

export TENCENTCLOUD_SECRET_ID=your_tencentcloud_secret_id
export TENCENTCLOUD_SECRET_KEY=your_tencentcloud_secret_key

Note

Please ensure that you securely manage your API keys. Do not expose them in public repositories or other public forums.

API Providers

Exif AI is designed to leverage various API providers for generating image descriptions and tags. We currently support five prominent providers, each offering unique capabilities and integration options. Below is a summary of the supported providers, including details on their requirements and features.

Supported Providers

• ZhipuAI: A cutting-edge AI service provider known for its advanced algorithms. Access to this service requires an API key.
• Ollama: An innovative local AI service that operates directly on your machine. This option does not require an API key, offering a seamless and private experience.
• Google Gemini: A robust AI service powered by Google, renowned for its high-quality image processing capabilities.
• Coze: Coze is a state-of-the-art AI platform designed for everyone, enabling the development of next-generation applications and chatbots.
• OpenAI: A preeminent AI service provider, recognized for its wide range of AI-powered tools and applications. Like ZhipuAI, it requires an API key for access.

Custom Providers

For users seeking to extend the functionality of Exif AI or integrate with additional AI services, we offer the flexibility to develop custom providers. By implementing the provider interface, you can create a custom provider to integrate with other AI services or tailor the image description generation process to your specific needs.

Configuration

OpenAI

To use OpenAI, you need to set the API key. You can do this by setting an environment variable:

export OPENAI_API_KEY=your_openai_api_key

If you wish to use a custom API service provider that is compatible with the OpenAI API, you can set the OPEN_API_BASEURL environment variable to point to the desired endpoint.

export OPEN_API_BASEURL=https://api.customprovider.com/v1

Zhipu AI

To use ZhipuAI, you need to set the API key. You can do this by setting an environment variable:

export ZHIPUAI_API_KEY=your_zhipuai_api_key

If you don't have a ZhipuAI account, you can sign up for one here.

Google Gemini

To use Google Gemini, you need to set the API key. You can do this by setting an environment variable:

export API_KEY=your_google_api_key

Coze Bot

To use Coze bot, you need to set the API key. You can do this by setting an environment variable:

export COZE_API_KEY=your_coze_api_key

When utilizing Coze within the Chinese region, configure the endpoint setting as follows:

export COZE_ENDPOINT=https://api.coze.cn

In Coze Bot, the model parameter corresponds to the bot's identifier. To interact with the Coze API, use the following command format:

exif-ai -a coze_bot -i image.jpg -m 7402199305639034921

Here, -a coze_bot specifies the Coze API, -i image.jpg indicates the input image file, and -m 7402199305639034921 is the unique identifier for the bot model you wish to use.

Ollama

Ollama runs locally and does not require an API key. Ensure that Ollama is installed and properly configured on your machine. Refer to the Ollama GitHub repository for installation and setup instructions.

To use remote Ollama service, you can defined the url in providerArgs:

exif-ai --providerArgs "http://ollama.example.com:8080" -a ollama -i image.jpg

providerArgs: ["http://ollama.example.com:8080"],

Develop

Prerequisites

• Node.js >=16
• pnpm

Clone the Repository

First, clone the Exif AI repository to your local machine:

git clone https://github.com/tychenjiajun/exif-ai.git
cd exif-ai

Install Dependencies

Next, install the required dependencies using pnpm.

pnpm install

Build

pnpm run build

Watch

pnpm run watch