jarvisit/HAM10000-classifier

---

language: en tags: - medical-imaging - dermatology - skin-disease - HAM10000 datasets: - HAM10000 metrics: - accuracy - precision - recall license: mit

HAM10000 Skin Disease Classifier

This model is trained on the HAM10000 dataset to classify 7 different types of skin lesions:

• Actinic keratoses (akiec)
• Basal cell carcinoma (bcc)
• Benign keratosis (bkl)
• Dermatofibroma (df)
• Melanoma (mel)
• Melanocytic nevi (nv)
• Vascular lesions (vasc)

Model Description

• Input: 224x224 RGB image
• Output: Probabilities for 7 skin lesion classes
• Architecture: [GUI]
• Training Dataset: HAM10000 (10,015 dermatoscopic images)

Performance

• Accuracy: [Accuracy değeri]
• Precision: [Precision değeri]
• Recall: [Recall değeri]

Usage

from transformers import AutoImageProcessor, AutoModelForImageClassification
import torch
from PIL import Image

processor = AutoImageProcessor.from_pretrained("jarvisit/HAM10000-classifier")
model = AutoModelForImageClassification.from_pretrained("jarvisit/HAM10000-classifier")

image = Image.open("path/to/image.jpg")
inputs = processor(image, return_tensors="pt")

with torch.no_grad():
    outputs = model(**inputs)
    predictions = outputs.logits.softmax(-1)

labels = model.config.id2label
predicted_label = labels[predictions.argmax(-1).item()]