ESP32-S3 AI Camera Module

The ESP32-S3 AI Camera combines a wide-angle IR camera, microphone and speaker with edge AI processing and Wi-Fi connectivity — ideal for smart monitoring, security, and voice-enabled IoT projects.

---

The ESP32-S3 AI Camera is an intelligent camera module built around the ESP32-S3 chip, designed for efficient video processing, edge AI tasks and voice interaction. It features a wide-angle infrared camera, along with an onboard microphone and speaker, making it well suited to applications such as electronic peepholes, baby monitors and licence plate recognition.

With built-in AI processing and Wi-Fi connectivity, the module supports edge-based image recognition as well as interaction with online AI models. It integrates easily into IoT systems, making it a practical choice for projects ranging from security and surveillance to smart devices and AI assistants.

Please note: the Gravity interface of V1.1 version has 3.3V output, so please do not connect to the input power supply.

Features

• Edge AI processing using the ESP32-S3 neural network engine
• Supports on-device image recognition with platforms including Edge Impulse, YOLOv5 and OpenCV
• Enables object detection and image classification without relying on the cloud
• Supports voice-controlled commands through ChatGPT integration
• Combines local AI processing with cloud-based model interaction for flexible IoT deployments
• Includes extensive documentation and example code, covering:
• Camera setup, video streaming and audio recording
• Image recognition and object classification
• OpenCV contour detection
• OpenAI integration for voice and image recognition
• Custom AI model training with Edge Impulse
• Integrated microphone and amplifier for:
• Voice recognition (ASR)
• Interactive dialogue using ChatGPT
• Voice-controlled automation and hands-free device management
• 160° wide-angle infrared camera with infrared illumination
• Built-in light sensor for automatic adaptation to lighting conditions
• Suitable for 24/7 monitoring, including low-light and evening use
• Ideal for baby monitoring, security surveillance and smart home systems
• Wireless connectivity via Wi-Fi and Bluetooth LE 5
• Supports remote video access and system control from mobile or connected devices
• Suitable for real-time monitoring, smart automation and remote surveillance applications
• Various AI capabilities
  • Edge image recognition (based on EdgeImpulse)
  • Online image recognition (OpenCV, YOLO)
  • Online large models for voice and image (ChatGPT)
• Equipped with a wide-angle night vision camera, infrared illumination, and all-day usability
• Onboard microphone and amplifier for voice interaction
• Offers a variety of AI models, with tutorial support for quick learning

Specifications

• Operating Voltage: 3.3V
• Type-C Input Voltage: 5V DC
• VIN Input Voltage: 3.7-15V DC
• Operating Temperature: -10~60°C
• Module Size: 42*42mm
• OV3660: 160° wide-angle infrared camera
• IR: Infrared illumination (IO47)
• MIC: I2S PDM microphone
• LED: Onboard LED (IO3)
• ALS: LTR-308 ambient light sensor
• ESP32-S3: ESP32-S3R8 chip
• SD: SD card slot
• Flash: 16MB Flash
• VIN: 3.7-15V DC input
• HM6245: Power chip
• Type-C: USB Type-C interface for power and code uploading
• Gravity
  • +: 3.3V
  • -: GND
  • 44: IO44/TX
  • 43: IO43/RX
• RST: Reset button
• BOOT: BOOT button (IO0)
• SPK: MX1.25-2P speaker interface
• MAX98357: I2S amplifier chip
• Camera Specifications
  • Sensor Model: OV3660
  • Pixels: 3 Megapixels
  • Sensitivity: Visible light, 940nm infrared
  • Field of View: 160°
  • Focal Length (EFL): 0.95
  • Aperture (F/No.): 2.0±5%
  • Distortion: ＜8%

Resources

• Wiki
• OpenAI Image Q&A Demo Code
• OpenAI Real-Time Communication/Demo for ESP-IDF

Package Contents

• 1x ESP32-S3 AI CAM development board (with camera)
• 1x Speaker
• 1x Gravity-4P I2C/UART sensor connection cable