Recognition Speed 0.3 s (Cold Start)
Recognition Distance 0.4 m–1.0 m
Liveness Detection TAR ≥ 98%

DM-SV001F1 High-Precision Binocular 3D Face Recognition Module for Smart Locks & Embedded Access Control

The DOMI DM-SV001F1 is a production-ready, ultra-compact dual-camera recognition module powered by a dual-core RISC-V SoC and dedicated NPU. Featuring real-time 3D liveness detection (anti-spoofing) with TAR ≥ 98% @ FAR = 10⁻⁶, it delivers millisecond-level authentication against 2D photos, video replays, and 3D silicone mask attacks.

Key Highlights

  • Sub-300ms cold boot and lightning-fast face matching.
  • Autonomous AI liveness detection blocking photos, screen replays, and 3D masks.
  • Automatically adapts to variations in lighting, facial expressions, hairstyles, and aging.
  • Local registration, group indexing, permission tiering, and batch deletion via UART.
Engineering support

Description

The DOMI DM-SV001F1 is an industrial-grade, mass-production-ready binocular 3D face recognition module engineered for ultra-compact embedded systems. Powered by a high-efficiency dual-core RISC-V processor and a dedicated hardware NPU, it embeds DOMI’s proprietary real-time liveness detection algorithm to deliver millisecond-level verification while preventing spoofing from high-resolution photos, curved screens, and 3D masks.

Designed specifically for battery-powered smart door locks and physical security terminals, the module balances ultra-low power consumption with financial-level anti-counterfeiting. The embedded AI engine executes complete image preprocessing, 3D geometric parallax verification, and face comparison locally on-chip, outputting encrypted authentication results via standard UART with zero dependency on external computing hosts.

Core System Capabilities:

  • Active NIR Binocular Architecture: Dual-channel synchronization for robust performance in complete darkness (0 Lux) and high ambient light.
  • Sub-300ms Cold Authentication: Instant wake-up to verification for seamless smart lock unlocking experiences.
  • Adaptive Self-Learning: Embedded algorithm dynamically updates facial profile features to adapt to makeup, aging, glasses, and expression changes.
  • Local User Base Management: On-chip face enrollment, group tiering, batch deletion, and database clearing managed entirely via UART commands.
Recognition Speed
0.3 s (Cold Start)
Recognition Distance
0.4 m–1.0 m
Liveness Detection
TAR ≥ 98%
Face Comparison
TAR ≥ 98% @ FAR = 10⁻⁶

Applications

The DM-SV001F1 binocular face recognition module delivers a high-security, contactless biometric alternative to legacy IC cards, optical fingerprints, and password systems. It is optimized for seamless embedded integration across multiple vertical markets:

🏠 Smart Living & Security

  • Smart electronic door locks & apartment entry
  • High-security electronic safes & vault boxes
  • Video intercom and smart doorbell panels

🏢 Access Control & Attendance

  • Pedestrian turnstiles and speed gates
  • Commercial access control systems
  • Enterprise time & attendance terminals

💳 Retail & FinTech Terminals

  • Smart POS machines & payment devices
  • Self-service retail & vending kiosks
  • VIP club & gym membership verification

🛡️ Identity & Public Sector

  • Medical insurance & healthcare identity kiosks
  • Driving school attendance verification
  • Industrial PC authorization & smart lockers

Specifications

ParameterSpecification
CPURISC-V Dual-Core Processor
NPUHigh-Performance NPU with Embedded AI Engine
ISP2-Lane ISP
RAM64 MB DDR3L
ROM16 MB SPI NOR Flash
Camera InterfaceMIPI CSI Interface
Other InterfacesUART (Serial Communication & Power Interface)
Power SupplyInput Voltage: 4 V ~ 9 V
Current Requirement: ≥ 500 mA
FOV (Field of View)54° (H) × 68° (V), 105° (D)
Face RecognitionRecognition Speed: 0.3 s (Cold Start)
Recognition Distance: 0.4 m ~ 1.0 m
Liveness Detection: TAR ≥ 98%
Face Comparison: TAR ≥ 98% @ FAR = 10⁻⁶
Image Sensor1/6″ CMOS Sensor
Resolution: 640 × 480 (VGA)
Interface: MIPI

FAQs

What anti-spoofing attacks can the DM-SV001F1 prevent?

The module features DOMI’s proprietary 3D binocular liveness detection, effectively preventing spoofing from high-resolution printed photos, 4K video replay on screens, curved paper masks, and 3D silicone head models.

What is the power consumption and cold-boot time for battery-operated locks?

The DM-SV001F1 supports a wide input voltage of 4V–9V and achieves a cold-start recognition speed of under 0.3 seconds, optimized for ultra-low-power sleep modes in smart door lock systems.

How does the module communicate with the host MCU?

It communicates via standard UART (TTL serial interface) using DOMI’s standardized hex command protocol for face enrollment, authentication, and user database management.

Need help integrating DM-SV001F1?

Get expert guidance from our sensing engineers on product selection, evaluation samples, system integration, and technical requirements for your application.

You May Also Like

DMAS2M001

View Details

DMOM2508CL

View Details

DM5005A

View Details

DMOM2501C

View Details