This model is a zero-shot object detection model capable of detecting target objects in images without specific category training. It is suitable for various application scenarios such as road damage detection.
Model Features
Zero-shot Object Detection
Detects target objects in images without requiring training data for specific categories.
Multilingual Support
Supports label inputs in French and English, suitable for multilingual environments.
Efficient Architecture
Based on the Swin-Tiny architecture, balancing detection accuracy and computational efficiency.
Model Capabilities
Zero-shot Object Detection
Multilingual Label Support
Image Analysis
Use Cases
Infrastructure Inspection
Road Damage Detection
Detects damage to infrastructure such as roads, curbs, and road signs.
Accurately identifies targets like damaged curbs and roads.
General Object Detection
Everyday Object Detection
Detects common objects in images, such as cats and remote controls.
Identifies and labels the positions of various everyday objects.
This repository is a fork of omlab/omdet-turbo-swin-tiny-hf and implements a custom task for zero-shot-object-detection on 🤗 Inference Endpoints. The code for the customized handler can be found in the handler.py.
✨ Features
Custom Task: Implements a custom task for zero - shot object detection on Hugging Face Inference Endpoints.
Easy Deployment: By selecting Custom as the task, you can use the handler.py file to deploy the model as an Inference Endpoint.
Dependency Management: The repository includes a requirements.txt file for downloading the timm library.
📦 Installation
The repository contains a requirements.txt file. You can use the following command to install the required libraries:
pip install -r requirements.txt
💻 Usage Examples
Basic Usage
To use this model as an Inference Endpoint, you need to select Custom as the task to use the handler.py file.
Expected Request Payload
{"inputs":{"image":"/9j/4AAQSkZJRgABAQEBLAEsAAD/2wBDAAMCAgICAgMC....",// base64 image as bytes"candidates":["broken curb","broken road","broken road sign","broken sidewalk"]}}
Run Request
The following is an example of how to run a request using Python and the requests library: