Add mold detection FastAPI backend

Dockerfile

@@ -0,0 +1,30 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy requirements first for better caching
+COPY requirements.txt .
+
+# Install Python dependencies
+RUN pip install --no-cache-dir --upgrade pip && \
+    pip install --no-cache-dir torch torchvision --index-url https://download.pytorch.org/whl/cpu && \
+    pip install --no-cache-dir -r requirements.txt
+
+# Copy application files
+COPY app.py model.py decision.py ./
+
+# Copy model file
+COPY resnet50_multitask_bio.pth ./
+
+# Expose port (HuggingFace Spaces uses port 7860)
+EXPOSE 7860
+
+# Run the application
+CMD uvicorn app:app --host 0.0.0.0 --port 7860
+
+

README.md

@@ -1,10 +1,82 @@
 ---
-title: Mold Detection Api
-emoji: 💻
+title: Mold Detection API
+emoji: 🦠
 colorFrom: blue
 colorTo: green
 sdk: docker
+app_file: app.py
 pinned: false
+license: mit
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Mold Detection API
+
+FastAPI backend for mold detection using multi-task ResNet50 deep learning model, deployed with Docker.
+
+# Mold Detection API
+
+FastAPI backend for mold detection using multi-task ResNet50 deep learning model.
+
+## Features
+
+- **Multi-task Learning**: Classifies mold types and detects biological material
+- **3-Level Decision System**: 
+  - High confidence (≥80%): "Mold"
+  - Medium confidence (50-80%) + biological detection: "Possible Mold"
+  - Low confidence: "Not Mold"
+- **RESTful API**: Easy integration with any frontend
+
+## API Endpoints
+
+### `GET /`
+Health check and API information
+
+### `GET /health`
+Simple health check
+
+### `POST /predict`
+Predict mold detection from an image
+
+**Request:**
+- Content-Type: `multipart/form-data`
+- File: Image file (jpg, png, jpeg)
+
+**Response:**
+```json
+{
+  "decision": "Mold" | "Possible Mold" | "Not Mold",
+  "mold_probability": 0.0-1.0,
+  "biological_probability": 0.0-1.0
+}
+```
+
+## Usage
+
+### Using curl:
+```bash
+curl -X POST "https://YOUR_USERNAME-SPACE_NAME.hf.space/predict" \
+  -F "file=@/path/to/your/image.jpg"
+```
+
+### Using Python:
+```python
+import requests
+
+url = "https://YOUR_USERNAME-SPACE_NAME.hf.space/predict"
+with open("test_image.jpg", "rb") as f:
+    response = requests.post(url, files={"file": f})
+    print(response.json())
+```
+
+## Documentation
+
+Interactive API documentation available at `/docs` endpoint.
+
+## Model
+
+- **Architecture**: ResNet50 with multi-task heads
+- **Input**: RGB images (224x224)
+- **Output**: 
+  - Classification head: 9 classes (mold class at index 4)
+  - Biological detection head: 2 classes (binary)
+

app.py

@@ -0,0 +1,68 @@
+from fastapi import FastAPI, UploadFile
+from fastapi.middleware.cors import CORSMiddleware
+from PIL import Image
+import torch
+import io
+from pathlib import Path
+from torchvision import transforms
+
+from model import MultiTaskResNet50
+from decision import final_decision
+
+app = FastAPI(
+    title="Mold Detection API",
+    description="FastAPI backend for mold detection using multi-task ResNet50",
+    version="1.0.0"
+)
+
+# Add CORS middleware for frontend
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],  # In production, replace with specific frontend URL
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+
+# Model path for HuggingFace Spaces (flat structure)
+model_path = Path("resnet50_multitask_bio.pth")
+
+print(f"Loading model from: {model_path.absolute()}")
+print(f"Model exists: {model_path.exists()}")
+
+model = MultiTaskResNet50()
+model.load_state_dict(torch.load(str(model_path), map_location=device))
+model.eval().to(device)
+print("✅ Model loaded successfully")
+
+transform = transforms.Compose([
+    transforms.Resize((224,224)),
+    transforms.ToTensor(),
+    transforms.Normalize(
+        mean=[0.485,0.456,0.406],
+        std=[0.229,0.224,0.225]
+    )
+])
+
+@app.get("/")
+async def root():
+    return {
+        "status": "healthy",
+        "message": "Mold Detection API is running",
+        "endpoint": "/predict",
+        "method": "POST",
+        "docs": "/docs"
+    }
+
+@app.get("/health")
+async def health():
+    return {"status": "healthy"}
+
+@app.post("/predict")
+async def predict(file: UploadFile):
+    image_bytes = await file.read()
+    img = Image.open(io.BytesIO(image_bytes)).convert("RGB")
+    img_tensor = transform(img).to(device)
+    return final_decision(model, img_tensor)

decision.py

@@ -0,0 +1,29 @@
+import torch
+import torch.nn.functional as F
+
+MOLD_HIGH_CONF = 0.80
+MOLD_LOW_CONF  = 0.50
+BIO_CONF       = 0.60
+
+def final_decision(model, img_tensor, mold_idx=4):
+    with torch.no_grad():
+        out = model(img_tensor.unsqueeze(0))
+        class_probs = F.softmax(out["class"], dim=1)[0]
+        bio_probs = F.softmax(out["bio"], dim=1)[0]
+
+    mold_p = class_probs[mold_idx].item()
+    bio_p  = bio_probs[1].item()
+
+    if mold_p >= MOLD_HIGH_CONF:
+        decision = "Mold"
+    elif mold_p >= MOLD_LOW_CONF and bio_p >= BIO_CONF:
+        decision = "Possible Mold"
+    else:
+        decision = "Not Mold"
+
+    return {
+        "decision": decision,
+        "mold_probability": round(mold_p, 3),
+        "biological_probability": round(bio_p, 3)
+    }
+

model.py

@@ -0,0 +1,20 @@
+import torch
+import torch.nn as nn
+from torchvision import models
+
+class MultiTaskResNet50(nn.Module):
+    def __init__(self, num_classes=9):
+        super().__init__()
+        self.backbone = models.resnet50(weights=None)
+        feat_dim = self.backbone.fc.in_features
+        self.backbone.fc = nn.Identity()
+        self.class_head = nn.Linear(feat_dim, num_classes)
+        self.bio_head = nn.Linear(feat_dim, 2)
+
+    def forward(self, x):
+        feats = self.backbone(x)
+        return {
+            "class": self.class_head(feats),
+            "bio": self.bio_head(feats)
+        }
+

requirements.txt

@@ -0,0 +1,9 @@
+torch
+torchvision
+fastapi
+uvicorn
+pillow
+numpy<2
+python-multipart
+
+