语言:

• 西班牙语
• 英语
• 多语言 许可证: 其他 标签:
• 图像分类 管道标签: 图像分类 微件示例:
• 图片链接: https://upserve.com/media/sites/2/Bill-from-Mezcalero-in-Washington-D.C.-photo-by-Alfredo-Solis-1-e1507226752437.jpg 示例标题: 收据
• 图片链接: https://templates.invoicehome.com/invoice-template-us-neat-750px.png 示例标题: 发票

InvoiceReceiptClassifier_LayoutLMv3 是一个基于LayoutLMv3微调的模型，用于将文档分类为发票或收据。

快速开始：使用原始模型

from transformers import (
    AutoModelForSequenceClassification,
    AutoProcessor,
)
from PIL import Image
from urllib.request import urlopen
model = AutoModelForSequenceClassification.from_pretrained("fedihch/InvoiceReceiptClassifier_LayoutLMv3")
processor = AutoProcessor.from_pretrained("fedihch/InvoiceReceiptClassifier_LayoutLMv3")
input_img_url = "https://upload.wikimedia.org/wikipedia/commons/thumb/0/0b/ReceiptSwiss.jpg/1024px-ReceiptSwiss.jpg"
with urlopen(input_img_url) as testImage:
    input_img = Image.open(testImage).convert("RGB")
    
encoded_inputs = processor(input_img, padding="max_length", return_tensors="pt")
outputs = model(**encoded_inputs)
logits = outputs.logits
predicted_class_idx = logits.argmax(-1).item()
id2label = {0: "发票", 1: "收据"}
print(id2label[predicted_class_idx])