File size: 3,940 Bytes
d39ab5e d203b52 d39ab5e 3a97deb e54264a d39ab5e e54264a fb7fa7e e54264a d39ab5e e54264a d39ab5e e54264a 2b74287 fcb4fc2 2b74287 d203b52 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 |
---
license: apache-2.0
language:
- en
base_model:
- google/siglip2-base-patch16-224
pipeline_tag: image-classification
library_name: transformers
tags:
- Waste
- Classification
---
# Augmented-Waste-Classifier-SigLIP2
> **Augmented-Waste-Classifier-SigLIP2** is an image classification vision-language encoder model fine-tuned from **google/siglip2-base-patch16-224** for a single-label classification task. It is designed to classify waste types using the **SiglipForImageClassification** architecture.
```py
Classification Report:
precision recall f1-score support
Battery 0.9987 0.9987 0.9987 3840
Biological 0.9998 0.9960 0.9979 4036
Cardboard 0.9956 0.9909 0.9932 3628
Clothes 0.9957 0.9914 0.9935 5336
Glass 0.9800 0.9914 0.9856 4048
Metal 0.9892 0.9965 0.9929 3136
Paper 0.9937 0.9891 0.9914 4308
Plastic 0.9865 0.9798 0.9831 3568
Shoes 0.9876 0.9990 0.9933 3990
Trash 1.0000 0.9939 0.9970 2796
accuracy 0.9926 38686
macro avg 0.9927 0.9927 0.9927 38686
weighted avg 0.9926 0.9926 0.9926 38686
```
The model categorizes images into 10 waste classes:
Class 0: "Battery"
Class 1: "Biological"
Class 2: "Cardboard"
Class 3: "Clothes"
Class 4: "Glass"
Class 5: "Metal"
Class 6: "Paper"
Class 7: "Plastic"
Class 8: "Shoes"
Class 9: "Trash"
```python
!pip install -q transformers torch pillow gradio
```
```python
import gradio as gr
from transformers import AutoImageProcessor
from transformers import SiglipForImageClassification
from transformers.image_utils import load_image
from PIL import Image
import torch
# Load model and processor
model_name = "prithivMLmods/Augmented-Waste-Classifier-SigLIP2"
model = SiglipForImageClassification.from_pretrained(model_name)
processor = AutoImageProcessor.from_pretrained(model_name)
def waste_classification(image):
"""Predicts waste classification for an image."""
image = Image.fromarray(image).convert("RGB")
inputs = processor(images=image, return_tensors="pt")
with torch.no_grad():
outputs = model(**inputs)
logits = outputs.logits
probs = torch.nn.functional.softmax(logits, dim=1).squeeze().tolist()
labels = {
"0": "Battery", "1": "Biological", "2": "Cardboard", "3": "Clothes",
"4": "Glass", "5": "Metal", "6": "Paper", "7": "Plastic",
"8": "Shoes", "9": "Trash"
}
predictions = {labels[str(i)]: round(probs[i], 3) for i in range(len(probs))}
return predictions
# Create Gradio interface
iface = gr.Interface(
fn=waste_classification,
inputs=gr.Image(type="numpy"),
outputs=gr.Label(label="Prediction Scores"),
title="Augmented Waste Classification",
description="Upload an image to classify the type of waste."
)
# Launch the app
if __name__ == "__main__":
iface.launch()
```
# Intended Use:
The **Augmented-Waste-Classifier-SigLIP2** model is designed to classify different types of waste based on images. Potential use cases include:
- **Waste Management:** Identifying and categorizing waste materials for proper disposal.
- **Recycling Assistance:** Helping users determine recyclable materials.
- **Environmental Monitoring:** Automating waste classification for smart cities.
- **AI-Powered Sustainability Solutions:** Supporting AI-based waste sorting systems to improve recycling efficiency. |