Fast Image decomposition API

This Python API for the Furn-E web app takes a generated interior design image and runs it through an object detection model to identify furniture in the image. It then performs a reverse image search with Microsoft Bing's image search API to display shopping results. Meant to be used with our Vue frontend.

Powered by Modal.com for parallel processing on-demand

1. To spin up a Modal instance, create a Modal account and get your API key.

   • Run this command to install modal client and generate token.

     pip install modal-client
     modal token new

     • The first command will install the Modal client library on your computer, along with its dependencies.

     • The second command creates an API token by authenticating through your web browser. It will open a new tab, but you can close it when you are done.

2. Deploy your modal project with the following command.

   modal deploy img_decomposition.main

How to use the API

1. Here's how you would upload your image file to the API using CURL.

curl -X POST "https://chriscarrollsmith--image-decompisition-api-v2-fastapi-app.modal.run/api/decompose" \
   -H "accept: application/json" \
   -H "Content-Type: multipart/form-data" \
   -F "image=@/path_to_your_image/image.jpg"

Alternatively, use the Javascript requests library:

// Select your input type=file and the image for upload
const fileInput = document.querySelector('input[type="file"]');
const imageData = fileInput.files[0]; // get the file

// Create a FormData instance and append the file
const formData = new FormData();
formData.append('image', imageData); // 'image' is the key expected by the backend

// Define the URL to your endpoint
const url = 'https://chriscarrollsmith--image-decomposition-api-v2-fastapi-app.modal.run/api/decompose';

// Make the POST request
fetch(url, {
method: 'POST',
body: formData // FormData will be sent as 'multipart/form-data'
})
.then(response => {
if (response.ok) {
   return response.json(); // if the response is good, get the JSON from the response
}
throw new Error('Network response was not ok.'); // if the response is not good, throw an error
})
.then(data => {
console.log(data); // Handle the data from the response
})
.catch(error => {
console.error('There has been a problem with your fetch operation:', error);
});

Sample response:

[
   {
      "uuid" : "1",
      "img" : "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mP8z/C/HgAGgwJ/lK3Q6wAAAABJRU5ErkJggg==",
      "keyword" : "pixel",
      "linklist" : [
            {
               "thumbnailURL" : "https://www.google.com/",
               "hostPageUrl" : "https://www.google.com/"
            }
         ],
      "description" : "Image of a single transparent pixel"
   }
]

Testing

Unit tests are built on the pytest framework. To run the tests, run the following command from the root directory of the project.

pytest tests