GOOGLE LLC (20240265731). Systems and Methods for On-Device Person Recognition and Provision of Intelligent Alerts simplified abstract
Contents
Systems and Methods for On-Device Person Recognition and Provision of Intelligent Alerts
Organization Name
Inventor(s)
Mohammad Afshar of Mountain View CA (US)
Saajan Shridhar of Mountain View CA (US)
George Alban Heitz, Iii of Mountain View CA (US)
Andrew C. Gallagher of Mountain View CA (US)
Michael C. Nechyba of Pittsburgh PA (US)
Joseph E. Roth of Longmont CO (US)
Systems and Methods for On-Device Person Recognition and Provision of Intelligent Alerts - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240265731 titled 'Systems and Methods for On-Device Person Recognition and Provision of Intelligent Alerts
The present document describes systems and methods for on-device person recognition and provision of intelligent alerts. The system includes a decentralized multi-camera system for on-device facial recognition. A device (e.g., security camera, video doorbell) captures images/video of a person, processes input image frames, detects face images, filters static faces, and aligns a rotation of the face to be upright and frontal. The device then filters low-quality face images and/or images having a large portion of the face occluded. The device computes a face embedding, compares it against a set of locally stored reference embeddings, and sends matching results to cloud services, which, based on the matching result, notifies the device owner whether the observed person is a known person or a stranger. Face detection and recognition computations are performed on device, not at the cloud. No sensitive information is transmitted off device and privacy is thus preserved.
- Decentralized multi-camera system for on-device facial recognition
- Processing of input image frames to detect and align face images
- Filtering of low-quality face images and images with occluded faces
- Comparison of face embeddings with locally stored reference embeddings
- Notification to device owner based on matching results
- On-device face detection and recognition computations for privacy preservation
Potential Applications: - Home security systems - Access control systems - Visitor management systems
Problems Solved: - Privacy concerns with cloud-based facial recognition - Efficient and accurate on-device person recognition - Real-time alerts for device owners
Benefits: - Enhanced privacy protection - Quick and reliable person recognition - Seamless integration with existing security systems
Commercial Applications: Title: On-Device Facial Recognition System for Enhanced Security This technology can be used in various commercial settings such as: - Residential security systems - Commercial buildings - Public spaces with access control needs
Prior Art: Readers can explore prior art related to on-device facial recognition systems in the field of computer vision and artificial intelligence research.
Frequently Updated Research: Researchers are constantly improving on-device facial recognition algorithms to enhance accuracy and speed in person recognition.
Questions about On-Device Facial Recognition System: 1. How does on-device facial recognition differ from cloud-based facial recognition systems? On-device facial recognition processes all computations locally, preserving privacy and reducing reliance on cloud services.
2. What are the key advantages of using a decentralized multi-camera system for facial recognition? A decentralized multi-camera system allows for more comprehensive coverage and accurate person recognition across different angles and lighting conditions.
Original Abstract Submitted
the present document describes systems and methods for on-device person recognition and provision of intelligent alerts. the system includes a decentralized multi-camera system for on-device facial recognition. a device (e.g., security camera, video doorbell) captures images/video of a person, processes input image frames, detects face images, filters static faces, and aligns a rotation of the face to be upright and frontal. the device then filters low-quality face images and/or images having a large portion of the face occluded. the device computes a face embedding, compares it against a set of locally stored reference embeddings, and sends matching results to cloud services, which, based on the matching result, notifies the device owner whether the observed person is a known person or a stranger. face detection and recognition computations are performed on device, not at the cloud. no sensitive information is transmitted off device and privacy is thus preserved.
