Detect objects in your images
Identify the top 3 objects in an image
Identify objects in your image
Upload an image to detect objects
Ultralytics YOLO11 Gradio Application for Testing
Detect traffic signs in uploaded images
Identify segments in an image using a Detectron2 model
Identify objects using your webcam
Detect marine vessels in images
Identify objects in an image
Identify objects in images using text queries
Identify and label objects in images using YOLO models
Upload image to detect objects
Transformers.js is a JavaScript library designed for object detection in images. It leverages the power of TensorFlow.js to provide an easy-to-use interface for detecting objects within images directly in the browser. Transformers.js enables developers to integrate robust object detection capabilities into web applications without requiring extensive machine learning expertise.
• Real-time Object Detection: Detect objects in images with high accuracy using pre-trained models.
• Pre-trained Models: Utilizes models like SSD MobileNet for efficient and accurate detection.
• Multiple Model Support: Easily switch between different object detection models based on your needs.
• Customizable Thresholds: Adjust detection confidence thresholds to filter results.
• Region of Interest (RoI): Focus on specific areas of an image for targeted detection.
• Image and Video Processing: Supports both image and video streams for versatile applications.
Install Transformers.js using npm or yarn:
npm install transformers.js
or
yarn add transformers.js
Import the library in your JavaScript file:
import { ObjectDetector } from 'transformers.js';
Initialize the detector with a pre-trained model:
const detector = new ObjectDetector('ssd_mobilenet_v2');
detector.load().then(() => console.log('Model loaded!'));
Load an image or video and pass it to the detector:
const img = new Image();
img.src = 'your-image.jpg';
img.onload = () => detector.detect(img).then(results => console.log(results));
Process the results to extract detection data (e.g., bounding boxes, class labels).
What browsers are supported by Transformers.js?
Transformers.js works on modern browsers that support TensorFlow.js and WebGL. This includes Chrome, Firefox, Safari, and Edge.
Can I use Transformers.js with video streams?
Yes! Transformers.js supports real-time object detection in video streams. You can pass a video element or a canvas to the detector for continuous detection.
How do I improve detection accuracy?
To improve accuracy, you can increase the confidence threshold or use a more accurate model like SSD MobileNet V2. Additionally, resizing images to the model's input size can enhance performance.