top of page
  • Writer's pictureJennifer Caballero

Introducing LibFaceRedaction: Empowering Privacy in Edge-Based Video Streaming and Surveillance

Updated: Apr 17

RidgeRun is dedicated to staying at the forefront of innovation and in sync with market requirements by crafting new products with cutting-edge Machine Learning solutions. In this blog post, we're thrilled to provide you with an early glimpse into our Face Redaction ongoing development effort: LibFaceRedaction.

Ensuring real-time privacy protection in edge-based video streaming and surveillance solutions is critical for some applications. LibFaceRedaction offers a comprehensive solution for concealing facial data information for safeguarding individuals identities. It harnesses the power of an ONNX machine learning model designed for face detection, ensuring both flexibility and performance.

Figure 1 provides an overview of the library's capabilities, which are tailored for deployment on an embedded system. The library's input is a video stream, typically sourced from a camera. Its primary functions include facial detection and the application of image effects to obscure faces within the video. Consequently, the library outputs the original video with concealed facial details.

Diagram showing the input/output of LibFaceRedaction. Video input from camera and video output with faces redacted.
Figure 1. LibFaceRedaction Overview

LibFaceRedaction is designed to support both CPU and GPU accelerated implementations, ensuring versatility.

Our ongoing commitment to enhancing this library underscores our dedication to meeting the evolving requirements of our customers. If you believe that LibFaceRedaction could serve as a valuable asset for enhancing privacy protection in your project, please don't hesitate to contact us.

Music by:, License code: ZIC94XFWSDLO6XYQ


Recent Posts

See All


bottom of page