Fine_tuned_Vit_trash_classification Open-source Image Classification Model - Precise Recognition for Efficient Garbage Classification

Home

Fine Tuned Vit Trash Classification

Developed by Aalaa

Image classification model based on ViT architecture, specifically fine-tuned for trash classification tasks

Image Classification

Transformers

Open Source License:Apache-2.0 #Fine-grained trash classification #ViT fine-tuned model #Few-shot learning

Downloads 46

Release Time : 1/3/2023

Model Overview

This Vision Transformer model was first pre-trained on ImageNet-21k, then fine-tuned on ImageNet 2012, and finally specifically fine-tuned for six-category trash classification tasks.

Model Features

Large-scale pre-training based

Pre-trained on ImageNet-21k (14 million images) with powerful feature extraction capabilities

Specialized trash classification optimization

Specifically fine-tuned for six categories: glass, paper, cardboard, plastic, metal, and trash

Transformer architecture

Adopts Vision Transformer architecture, processing images as sequences of 16x16 patches

Model Capabilities

Image classification

Trash identification

Material identification

Use Cases

Trash classification

Recyclable material identification

Identifies recyclable materials such as glass, paper, cardboard, plastic, and metal

Accurate classification of various recyclable materials

Automatic trash sorting

Used in smart trash bins or waste processing facilities for automatic sorting systems

Improves trash sorting efficiency and accuracy

🚀 Vision Transformer (base-sized model)

A Vision Transformer (ViT) model pre - trained on ImageNet - 21k and fine - tuned on ImageNet 2012 for image classification.

🚀 Quick Start

Vision Transformer (ViT) model pre-trained on ImageNet-21k (14 million images, 21,843 classes) at resolution 224x224, and fine-tuned on ImageNet 2012 (1 million images, 1,000 classes) at resolution 224x224. It was introduced in the paper An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale by Dosovitskiy et al. and first released in this repository. However, the weights were converted from the timm repository by Ross Wightman, who already converted the weights from JAX to PyTorch. Credits go to him.

Disclaimer: The team releasing ViT did not write a model card for this model so this model card has been written by the Hugging Face team.

✨ Features

Pre-trained on ImageNet-21k and fine-tuned on ImageNet 2012.
Applicable for image classification tasks.

📦 Dataset

The dataset used consists of six classes: glass, paper, cardboard, plastic, metal, and trash. Currently, the dataset consists of 2527 images:

501 glass
594 paper
403 cardboard
482 plastic
410 metal
137 trash

📚 Fine_tuned Notebook

This notebook outlines the steps from preparing the data in the Vit-acceptable format to training the model Notebook

💻 Usage Examples

Basic Usage

from transformers import AutoFeatureExtractor, AutoModelForImageClassification
from PIL import Image
import requests
url = 'https://www.estal.com/FitxersWeb/331958/estal_carroussel_wg_spirits_5.jpg'
image = Image.open(requests.get(url, stream=True).raw)

feature_extractor = AutoFeatureExtractor.from_pretrained("Aalaa/Fine_tuned_Vit_trash_classification")
model = AutoModelForImageClassification.from_pretrained("Aalaa/Fine_tuned_Vit_trash_classification")

inputs = feature_extractor(images=image, return_tensors="pt")
outputs = model(**inputs)
logits = outputs.logits

predicted_class_idx = logits.argmax(-1).item()
print("Predicted class:", model.config.id2label[predicted_class_idx])

For more code examples, we refer to the documentation.

📄 License

This project is licensed under the Apache-2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご