웨비나

Ultra-Compact and Low-Power Edge AI Design in Practice: ESP32-S3 x YOLO Nano Object Detection

2026.02.19 10:30~12:00 김봉조 대표 / e4ds

Is real-time video analysis possible at the edge without expensive GPUs and high-power SBCs?

This webinar is designed for embedded developers and AI engineers

covering implementation methods for Edge AI video recognition using ESP32-CAM based on ESP32-S3 and YOLO Nano.

Ultra-lightweight object detection models that can operate even in MCU environments and

through object and face recognition demos that can be verified immediately on actual boards,

we share practical strategies for directly connecting ideas to prototypes and services.

Hardware Insight – Understanding ESP32-CAM Architecture

✔️ Analysis of AI processing flow based on ESP32-S3

✔️ Understanding camera interface and memory structure

Model Optimization – YOLO Nano Lightweight Strategy

✔️ Model architecture design tailored to MCU environments

✔️ Quantization application methods and performance optimization points

Real-world Demo – Real-time Object and Face Recognition Implementation

✔️ Edge AI demo operating on a single board

✔️ Event-driven video analysis scenarios

Edge AI Strategy – Low-Power, Cloud-Free Design

✔️ On-site processing structure without cloud dependency

✔️ System design strategy from latency, power consumption, and security perspectives

✔️ Embedded developers who want to add advanced AI functionality to ESP32-based projects

✔️ Startup personnel seeking to implement low-cost Edge AI prototypes without expensive equipment

✔️ AI engineers curious about deep learning model optimization and hardware porting processes

✔️ Technology leaders looking for practical, hands-on techniques combining IoT, vision, and AI

✔️ Webinar presentation materials provided in PDF

– Hands-on guide based on Arduino IDE / ESP-IDF / VS Code

Object and Face Recognition Starting with a Single ESP32-CAM Board

Through this webinar, verify practical implementation methods for On-Device AI firsthand.