WiMi Develops Holographic Face Recognition AI Chip System

The face recognition acquisition and analysis system, mainly through the cloud server cluster, the traditional CPU, and the master chip, calculates power shortage and energy cost high, real-time acquisition and recognition, and immediate reaction effect, causing some critical data information omission, or error detection, deployment cost is very high. WIMI Hologram Cloud, Inc. (NASDAQ: WIMI) develops a holographic face recognition AI chip system based on edge computing, putting recognition, collection, and analysis at the terminal. On the one hand, it effectively improves algorithm computing power optimization; on the other hand, a private domain can be established to effectively protect data security. It can be used to control some key departments, key enterprises, or high-end factories and the security management of industrial parks, office buildings, and apartments, with convenient, safe, and efficient deployment.

The holographic face recognition AI chip system, based on edge computing, because it is edge computing, is different from the traditional face recognition and human/card comparison methods. It can realize synchronous face-tracking acquisition and face attribute analysis feedback results. It can obtain the holographic high-density face data of the collected person at the front end: face properties, appearance, characteristics, collection time, geographical location, and other essential information, and it can identify and distinguish the features of the collected person. The holographic face recognition AI chip system from WiMi can be based on the existing security video system, combined with edge computing technology, AI calculates force acceleration technology, deep learning algorithm, holographic data gain technology, convolutional neural network technology, face recognition and acquisition technology, formed the based on edge computing holographic face recognition AI chip system, merge the original security system upgrade.

WiMi’s Holographic face recognition AI chip system based on edge computing can use a time window as a sampling period and can also be set according to specific target personnel, such as the on-job time of crucial positions and geographical space. For example, multiple sampling is conducted in a sampling period, and the optimal sampling is selected as the final sampling information. All faces must be identified if multiple faces appear in a sampling frame. For the completely unrecognizable faces, ignore the retention strategy and focus on grasping again in the subsequent frame or other monitoring positions until recognition, to ensure comprehensive recognition, the integrity of the data in the management area of the system, and maintain the regional security. The holographic face recognition AI chip based on edge computing is used for high-density dynamic personnel information collection for surveillance videos in key places, providing structured primary data for safety management and production. The system includes a video access port and docking with the existing surveillance video.

The holographic image decodes the frame module, decodes the central control module instruction, and extracts the frame to be analyzed.

The holographic image optimization processing module analyzes the image acquisition in the extraction frame according to the instructions of the central control module, accelerates the holographic image optimization of the sampled structure, and feeds back to the central control module. The new information obtained, such as the missing information, will continue to be issued by the main control module.

The edge computing and algorithm acceleration module, including the core computing unit, is based on ARM architecture, embedded multi-layer CNN convolutional neural network, algorithm operation on low power consumption, and high parallel computing performance.

Face collection and analysis module, collect faces for identification and segmentation, analyze face photos, geographic information, time information, collected face information, gender, age, race, mask, glasses, etc.

The central control module realizes the management of the sampling process and the comprehensive control and management of other modules.

The data storage and notification module stores the collected personnel information locally and notifies the external system according to the information level.

WiMi’s holographic face recognition AI chip system is based on edge computing, the front-end through the edge computing dynamic holographic face recognition algorithm, through video access video holographic decoding, and the face of the image, tracking, capture, heavy, and characteristic value as information identification, build past personnel information, complete personnel information collection, and can realize private domain management, improve information security level. In addition, the system deployment is convenient, can be external, rack, mobile a variety of deployment ways, can connect this device directly to the existing high-definition network camera, directly in the front can complete part of the video structured work, get high quality face structured data, and improve the back-end intelligent identification and analysis speed and calculation efficiency, make full use of the existing stock camera, through the form of the plug-in can directly to the current unstructured video upgrade to intelligent structured data.

WiMi Hologram Cloud, Inc. (NASDAQ: WIMI) holographic face recognition AI chip system based on edge computing can conduct high-density dynamic holographic face collection in a complex environment, which can be used in various important occasions, as well as digital camera and intelligent front end for face information collection, to meet different needs of security information collection.