inflearn logo
inflearn logo

[Practical AIoT] Perfect Preparation for Smart Mirror Makerthon: LLM, CV, and Hardware Design

Solve the point where 80% get stuck at makeathons. Complete Raspberry Pi · Computer Vision · LLM · 3D Design in 4 weeks! Achieve top rankings at makeathons with a demonstrable smart mirror PoC!

(4.8) 5 reviews

23 learners

Level Basic

Course period Unlimited

Python
Python
Raspberry Pi
Raspberry Pi
Arduino
Arduino
AI
AI
LLM
LLM
Python
Python
Raspberry Pi
Raspberry Pi
Arduino
Arduino
AI
AI
LLM
LLM
날개 달린 동전

Recommend Course to grow and earn commission!

날개 달린 동전

Marketing Partners

Recommend Course to grow and earn commission!

What you will gain after the course

  • "Integrated PoC Implementation Capability"

  • LLM API Practical Template (Reusable Framework)

  • MediaPipe FaceMesh-based "Recognition UX PoC"

  • Raspberry Pi ↔ Arduino Serial Communication (Python) Capability

  • Assemblable smart mirror hardware design deliverables

This course is a practical preparatory training designed to help you complete the core competencies most frequently required in Smart Mirror Makerthon/AIoT Makerthon in the form of "pre-mission assignments." Rather than a fragmented list of technologies, it is structured with the goal of creating an integrated PoC (Proof of Concept) that can actually be demonstrated.

Students will complete the following 4 essential deliverables through the course.

  1. LLM API-Based Conversation Response Module

  • Going beyond simple LLM API calls, we create a "production-ready template" equipped with role separation (system/user), conversation history maintenance, retry/timeout, and logging.

  • We'll also organize a response format that outputs short, card-style summary responses (at-a-glance format) suitable for smart mirrors.

  1. MediaPipe FaceMesh Facial Landmark + Filter PoC

  • Extract FaceMesh landmark coordinates and implement a computer vision demo that tracks 1 filter (sunglasses/mask/sticker, etc.) on the face in real-time.

  • Establish a foundation that can be expanded to "user recognition-based UI/interaction" in the future.

  1. Raspberry Pi ↔ Arduino Serial Communication (Python) for Sensor Data Transfer

  • The sensor values read by Arduino are transmitted via serial (UART), and received, parsed, and exception-handled by Python (pyserial) on the Raspberry Pi.

  • We've structured this to proactively solve communication/permission/data corruption/timeout issues, which are the most common bottlenecks in makerthons.


Additionally, in the optional section, you can expand on the following based on your team's capabilities and goals.

  • Weather data input on Raspberry Pi (including city change/refresh)

  • Firebase or Supabase DB Integration (storing/retrieving values)

  • Responsive Web UI (Mobile/Desktop)

  • AWS Deployment Basics

Upon completing the course, students will not just have "a little knowledge of each technology," but will start with an integrated demo ready for immediate demonstration at a makeathon + submission documents/evidence (logs, videos, README).

Recommended for
these people

Who is this course right for?

  • Prospective makerthon/hackathon participants

  • Developers/students who want to create projects with Raspberry Pi and Arduino

  • People who want to integrate generative AI (LLM) based features into actual products/projects

  • Someone who needs a Computer Vision (Face Recognition) PoC

  • For those who need a portfolio that includes hardware design

  • Team Project Leader/Educator (Instructor/Teaching Assistant)

Need to know before starting?

  • Basic Programming

Hello
This is kodekorea

23

Learners

5

Reviews

4.8

Rating

1

Course

We are KodeKorea, a company specializing in IT education.

You can learn the latest IT technologies in various fields faster than anyone else.

We focus on bottom-up education covering everything from simple coding to advanced technology stacks and hackathon competitions.

You can complete it compactly and in a short period with maximum efficiency.

🔗Homepage |

🧑‍💻Instagram | @kodekorea

📭Email | shain1912@naver.com

More

Curriculum

All

22 lectures ∙ (8hr 49min)

Course Materials:

Lecture resources
Published: 
Last updated: 

Reviews

All

5 reviews

4.8

5 reviews

  • guswls101613772님의 프로필 이미지
    guswls101613772

    Reviews 1

    Average Rating 5.0

    5

    43% enrolled

    • ksksyy259208님의 프로필 이미지
      ksksyy259208

      Reviews 2

      Average Rating 5.0

      5

      43% enrolled

      • m2khk024921님의 프로필 이미지
        m2khk024921

        Reviews 1

        Average Rating 5.0

        5

        100% enrolled

        • tkddms127285님의 프로필 이미지
          tkddms127285

          Reviews 4

          Average Rating 3.5

          4

          100% enrolled

          It was an informative lecture that allowed even a beginner who knows nothing to understand everything, as long as they silently and steadily keep up with the course.

          • 64kmqkscks5943님의 프로필 이미지
            64kmqkscks5943

            Reviews 1

            Average Rating 5.0

            5

            100% enrolled

            Similar courses

            Explore other courses in the same field!

            $83.60