Upload 3 files

Dockerfile

@@ -0,0 +1,14 @@
+# Use the official Python 3.10.9 image
+FROM python:3.10.9
+
+# Copy the current directory contents into the container at .
+COPY . .
+
+# Set the working directory to /
+WORKDIR /
+
+# Install requirements.txt 
+RUN pip install --no-cache-dir --upgrade -r /requirements.txt
+
+# Start the FastAPI app on port 7860, the default port expected by Spaces
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,116 @@
+from pydantic import BaseModel, Field, HttpUrl, ConfigDict
+from typing import List
+from openai import AsyncOpenAI
+from fastapi import FastAPI, HTTPException
+
+class BusinessCardResponse(BaseModel):
+    is_valid_business_card: bool = Field(title = "Is This a Valid Business Card", description="To flag whether the user inputted image is a business card")
+    name: str
+    job: str
+    company: str
+    phone_number: List[int]
+    email: List[str]
+    address: List[str]
+    website: List[str]
+
+client = AsyncOpenAI(api_key = 'sk-proj-BszN5uaYsFxdGaqnrw6cT3BlbkFJmSFEl0vKfMZJYOAePJ26')
+
+app = FastAPI(title="Business Card Scanner API", description = "A FastAPI app to do physical business card scanning & information extraction", version = "0.1.0")
+
+class ImageRequest(BaseModel):
+    url: str
+
+    model_config = ConfigDict(
+        json_schema_extra={
+            "examples": [
+                {
+                    "url" : "https://marketplace.canva.com/EAFIF52_A0Y/1/0/1600w/canva-putih-biru-profesional-minimalis-kartu-nama-bisnis-AmVxO2cXfvA.jpg"
+                }
+            ]
+        }
+    )
+
+prompt = """
+    ### CONTEXT ###
+    I want to create a physical-to-digital business card scanner application. This app allows the user to take a picture of a physical business card, extract the information inside the picture, and create a digital business card.
+
+    ### OBJECTIVE ###
+    The user will give you an image of a business card. Your first task is to check whether it is a business card. If it is a business card, you must extract these information from the card:
+    1. The cardholder's name (including the salutation & degrees, if any)
+    2. Job/occupation of the cardholder
+    3. The company name
+    2. Phone Number(s) (list of integer, remove plus/+ sign and any other symbols)
+    3. Email address (list)
+    4. Home/office address (list)
+    5. Personal/company Website (list)
+
+    ### OUTPUT ###
+    You must only return a JSON output in a snippet with this schema:
+    ```json
+    {
+        "is_valid_business_card": bool,
+        "name": str,
+        "job": str,
+        "company": str,
+        "phone_number": List[int],
+        "email": List[str],
+        "address": List[str],
+        "website": List[str]
+    }
+    ```
+    You must only return the JSON output only, without any additional comments!
+
+    ### SAMPLE RESPONSE ###
+    ```json
+    {
+        "is_valid_business_card": True,
+        "name": "Matthew Farant, BSc.",
+        "job": "Business Intelligence Analyst",
+        "company": "TipTip",
+        "phone_number": [62812345678, 62312345678],
+        "email":["[email protected]", "[email protected]"],
+        "address":["Jalan Kepang Lima No. 8"],
+        "website":["www.matthew.com"]
+    }
+"""
+
+@app.post("/api/scan", response_model=BusinessCardResponse)
+async def extract_card(image_request: ImageRequest) -> BusinessCardResponse:
+    response = await client.chat.completions.create(
+        model="gpt-4o-mini",
+        messages=[
+            {
+                "role": "system", 
+                "content": [
+                    {
+                        "text": prompt,
+                        "type": "text"
+                    }
+                ]
+            },
+            {
+                "role": "user",
+                "content": [
+                    {
+                        "type": "text",
+                        "text": """Extract the business card information from the image and put them in a JSON format"""
+                    },
+                    {
+                        "type": "image_url",
+                        "image_url": {"url": image_request.url}
+                    }
+                ]
+            },
+            {
+                "role": "assistant",
+                "content": "```json"
+            }
+        ],
+        temperature=1,
+        max_tokens=300,
+        top_p=1,
+        frequency_penalty=0,
+        presence_penalty=0
+    )
+
+    return BusinessCardResponse.model_validate_json(response.choices[0].message.content.replace('```json','').replace('```','').strip())

requirements.txt

@@ -0,0 +1,3 @@
+fastapi==0.109.2
+pydantic==2.7.0
+openai==1.12.0