Chinese Food Classification with ResNet-50 (ChineseFoodNet)

Fine-tuned ResNet-50 for Chinese food classification using the ChineseFoodNet dataset. (Kaggle: ChineseFoodNet)

Model Details

• Base Model: ResNet-50
• Dataset: ChineseFoodNet (101,565 images)
• Classes: 114 Chinese food categories
• Test Accuracy: 75.4%

Performance

• Test Accuracy: 0.7536
• Macro F1-Score: 0.7409
• Weighted F1-Score: 0.7520

Usage

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

processor = AutoImageProcessor.from_pretrained("Albertbeta123/resnet-50-chinese-food")
model = ResNetForImageClassification.from_pretrained("Albertbeta123/resnet-50-chinese-food")

image = Image.open("chinese_food_image.jpg")
inputs = processor(image, return_tensors="pt")

with torch.no_grad():
    logits = model(**inputs).logits
    predicted_class_id = logits.argmax(-1).item()
    predicted_class = model.config.id2label[predicted_class_id]
    confidence = torch.nn.functional.softmax(logits, dim=-1).max().item()

print(f"Predicted: {predicted_class} ({confidence:.1%})")