FaceTron
Created13shivam
Provides a high-performance face embedding server utilizing ONNX Runtime for scalable face recognition applications.
About
FaceTron is a modular face recognition and embedding server built for high performance and scalability. Leveraging ONNX Runtime, it supports dynamic loading of multiple face recognition models, enabling offline deployment and adaptable performance across various environments. FaceTron offers an OpenAPI Specification (OAS) endpoint compatible with Model Connector Protocol (MCP) metadata and integrates with OpenTelemetry for comprehensive observability, making it ready for integration into model orchestration pipelines.
Key Features
- 0 GitHub stars
- Built-in OpenTelemetry tracing for observability.
- Dynamic multi-model support for face embedding.
- MCP metadata support for model information.
- Face detection and visualization with bounding boxes.
- OpenAPI 3.1 and Swagger UI for easy API integration.
Use Cases
- Building scalable face recognition systems.
- Offline face recognition model deployment.
- Integrating face embedding into existing applications via APIs.
