USB computer-on-module

EthernetWiFimachine learning

Introduction
The Edge AI SoC Module is a compact, high-performance computing module delivering on-device AI and advanced video processing for vision-based products. It provides flexible camera interfaces and multiple network options to accelerate development of security, industrial and smart home applications.

Key benefits

• Fast deployment: out-of-the-box operation to launch a first vision project in ~30 minutes
• Reduced development time: simplified connection and configuration tools can cut development cycles by up to 69%
• Lower hardware cost: supports multiple OS and development environments to reduce hardware expenses by up to ~30%
• High-performance edge AI: integrated AI acceleration engine for local inference
• Flexible camera connectivity: supports MIPI and DVP interfaces
• Driver-free UVC operation: works as a USB camera without drivers for easy integration
• Multiple network transmission options: Wi‑Fi and wired Ethernet for local and remote streaming

Applications

• Security surveillance and access control systems
• Smart home devices with embedded vision
• Industrial inspection and machine vision
• Other edge-vision scenarios (up to 8 supported application types)

Technical details
The module delivers efficient video encoding/decoding with low-latency transmission and is compatible with mainstream AI frameworks for stable, scalable edge intelligence. The integrated AI acceleration engine supports algorithms such as object detection, face recognition and real-time video analytics. Camera interfaces include MIPI and DVP, and the module supports both Wi‑Fi and wired Ethernet. USB Video Class (UVC) driver-free support enables plug-and-play use as a USB camera.

Specifications

• Product ID: 291935820
• Type: Edge AI SoC computing module
• Camera interfaces: MIPI, DVP
• AI: integrated AI acceleration engine; compatible with mainstream AI frameworks
• Video: efficient encoding/decoding, low-latency transmission
• Network: Wi‑Fi, wired Ethernet
• USB: UVC (driver-free USB Video Class)
• Algorithms supported: object detection, face recognition, real-time video analytics
• Deployment: out-of-the-box operation (~30 minutes to first project); simplified configuration
• Development benefits: reduce development cycles up to 69%; reduce hardware costs up to ~30%