Skip to content

Use custom trained Cloud Annotations models with TensorFlow.js

License

Notifications You must be signed in to change notification settings

cloud-annotations/javascript-sdk

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

44 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Cloud Annotations JavaScript SDK

NPM Version NPM Downloads

The Cloud Annotations SDK makes it easy to use your custom trained object detection or classification models in the browser or on the backend with TensorFlow.js. Simply load() your model_web folder and pass the loaded model a <canvas>, <img> or <video> reference.

Installation

npm install @cloud-annotations/models

Usage

Load a model

import models from '@cloud-annotations/models'

const model = await models.load('/model_web')

Object detection

const img = document.getElementById('img')
const predictions = await model.detect(img)

// predictions =>
[{
  label: 'dog',
  bbox: [x, y, width, height],
  score: 0.92
},
{
  label: 'cat',
  bbox: [x, y, width, height],
  score: 0.72
}]

Classification

const img = document.getElementById('img')
const predictions = await model.classify(img)

// predictions =>
[
  { label: 'dog', score: 0.92 },
  { label: 'cat', score: 0.72 }
]

Usage via Script Tag

No npm install required. Just import via the script tag.

<script src="https://cdn.jsdelivr.net/npm/@cloud-annotations/models"></script>
<script>
  const img = document.getElementById('img')
  models.load('/model_web')
    .then(model => model.detect(img))
    .then(predictions => {
      console.log(predictions)
    })
</script>

Example usage: Real-Time Object Detection With React.