faceverify

Deployment Guide

Deploy FaceVerify to production.

---

Docker Deployment

Dockerfile

FROM python:3.11-slim

WORKDIR /app

# Install system dependencies
RUN apt-get update && apt-get install -y \
    libgl1-mesa-glx \
    libglib2.0-0 \
    && rm -rf /var/lib/apt/lists/*

# Install Python dependencies
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

# Copy application
COPY . .

# Run
CMD ["python", "-m", "faceverify.api"]

Build and Run

docker build -t faceverify:latest .
docker run -p 8000:8000 faceverify:latest

---

Environment Configuration

Production Settings

# .env
FACEVERIFY_DETECTOR=retinaface
FACEVERIFY_THRESHOLD=0.70
FACEVERIFY_ENABLE_GPU=true
FACEVERIFY_BATCH_SIZE=16

---

Scaling

Horizontal Scaling

FaceVerify is stateless - scale horizontally with load balancer:

Load Balancer
    |
    +-- FaceVerify Instance 1
    +-- FaceVerify Instance 2
    +-- FaceVerify Instance 3

GPU Scaling

For high throughput:

1. Use GPU instances
2. Enable batching
3. Consider model serving (TensorFlow Serving)

---

Monitoring

Health Check Endpoint

@app.get("/health")
def health():
    return {
        "status": "healthy",
        "version": __version__,
    }

Metrics to Monitor

Metric	Description
Verification latency	p50, p95, p99 response times
Error rate	Failed verifications / total
Throughput	Verifications per second
Memory usage	RAM consumption
GPU utilization	If using GPU

---

Troubleshooting

High Memory Usage

• Reduce batch size
• Process images sequentially
• Use smaller embedding model

Slow Performance

• Enable GPU
• Use faster detector (opencv)
• Pre-compute embeddings

Model Loading Errors

[!TIP] First run downloads model weights (~90MB). Ensure:

• Internet connectivity
• Write access to cache directory
• Sufficient disk space

This site is open source. Improve this page.