홈 오토메이션과 게임을 위한 혁신적인 머신 러닝 프로젝트

이 영상에서는 몸의 움직임을 이용한 드론 제어, 선글라스를 이용한 홈 오토메이션, 게임용 에임봇과 같은 머신러닝 프로젝트에 대해 설명합니다.

### 혁신적인 머신 러닝 프로젝트 아이디어
* 🚁 몸의 움직임으로 드론 제어: **Murtaza는 PID 컨트롤러를 사용하여 신체 움직임을 드론 좌표에 매핑**하여 수동 입력 없이도 정밀하게 제어할 수 있는 방법을 시연합니다.
* 🕶 선글라스를 이용한 홈 오토메이션: Nick의 프로젝트는 선글라스에 부착된 카메라를 사용하여 **2계층 신경망으로 구동되는 제스처 인식**을 통해 가전제품을 제어합니다.
* 🎯 게임용 픽셀 스캐닝 에임봇: **2~3일 동안 훈련된 카말의 에임봇**은 화면 색상을 기반으로 목표물을 감지하여 게임에서 **조준 정확도를 크게 향상**시킵니다.
* 🌡 실시간 온도 측정: RGB 및 적적외선 카메라를 사용하여 물체의 온도를 측정하는 프로젝트로, **뜨거운 물체는 빨간색**으로, **차가운 물체는 파란색**으로 표시합니다.
* 🏹 양궁 봇은 모터와 목표물 식별을 위한 깊이 카메라를 사용하여 **활 각도, 줄 당기기, 화살 발사** 등을 제어합니다.
* 🧙 이 프로젝트는 지팡이로 특정 글자를 공중에 그리면 **머신 러닝을 통해 패턴을 인식**하여 상자를 여는 등의 동작을 트리거합니다.
* 🛒 빠른 쇼핑 결제를 위한 자동 청구: **카메라로 물체를 식별**하고 **로드셀로 무게를 측정**하여 자동으로 쇼핑 청구서를 생성하는 DIY 설정입니다.
* 🗑 쓰레기 분리수거를 위한 스마트 쓰레기통: **사무엘의 스마트 쓰레기통은 식별에 따라 다양한 종류의 쓰레기를 수거**하기 위해 회전하여 가정에서 적절한 쓰레기 관리를 도와줍니다.

hi everyone welcome back to our channel. In this video, we'll discuss some machine learning project ideas. So let's get started. Number 12: Operating a drone requires some skill and regular practice, but if you don't have that much time, then you could take some help from this project. Here, Murtaza shows how he controls a drone using his body movement. He used a PID controller that maps the body's position into coordinates. As he moves, the coordinates get changed from the neutral position, and accordingly, the drone follows the new coordinates. In this way, the drone always follows the body movements while maintaining a certain safe distance from the body. Number 11: No need to spend your hard-earned money on home automation devices because all you need is a microcontroller board and regular sunglasses to control various appliances at your home. Yup, you heard it right. Nick attached a camera at the front of his glasses that streams the video to the Jetson Nano. Using this, he just looks at the objects he wants to control, makes a gesture, and

voila, the task is completed. The objects and gestures are identified using a two-layer neural network, which triggers certain objects like opening and closing the lights. Number 10. If you've not got gaming skills and missed the targets most of the time, then this project can be a great hack. Kamal made a pixel scanning Aimbot that can easily detect targets based on the colors of the screen. In order to make it undetectable by the game software, he made a physical setup in which he added four omnidirectional wheels to his mouse. With this, the bot can move easily in any direction. He trained the model for two or three days, and after that, the Aimbot was aiming better than most of the pro players. Number nine. With the evolution of artificial intelligence, the way of solving problems has also evolved. One such project based on it is the way to measure temperature. It uses an RGB and infrared camera to measure the temperature of the objects in real time. On the left side of the screen, you could see the objects and their temperature.

using the RGB camera, while the infrared camera gives a visual representation of the object's hotness. The hot objects can be identified with red color, and the cold objects with blue color. Number eight, if you're into archery, then you might know how hard it is to hit a bullseye. It requires patience, dedication, and a lot of practice, but this Aimbot can do that pretty easily. The archery bot controls the bow angle of attack using Metal Gear Motors. A stepper motor pulls the string and arrow back, and a Servo releases the bow to shoot using an Intel RealSense depth camera. The color and depth of a Target are identified using this data. The initial angle and the velocity of the arrow are calculated. Altium is a PCB design software that's both powerful and easy to use. With its intuitive interface and comprehensive set of features, Altium Designer makes it easy to create high-quality PCB designs. And with all-new Altium 365, now you can share your PCB designs with anyone from anywhere with a single click. With Octopart, you can get real-time component

insights as you design your PCB, and Altium. Check the description for more details. Number seven in the above-discussed drone project. The Drone was controlled using body movement, and you can't do some quick actions using that. For example, flipping a drone like this. To make this happen, you can learn how to operate it using hand gestures. In this project, Murtaza added some safety features to control. Like the Drone can be operated only when a base is detected, and the left hand is used to make gestures. Also, the hand must be present inside a specific box. If these conditions are not met, the Drone will not operate. [Music] Number five. If you want to improve your playing style and check whether your form is good or bad, then this project could be a great help. Samuel attached a microcontroller to his table tennis bat, and with the motion trainer, he trained the model to identify different types of shots. It even tells whether you're playing a particular shot properly or not. It's impressive how he managed to place the components compactly over the bat.

that does not create any hindrance while playing. Number five, if you're a die-hard fan of the Harry Potter series and want to relive the experience, then we have something magical for you. When you draw certain letters in the air with this wand, it'll perform some activity. Like drawing the letter A represents alohomara, which opens up a box. The tip of the wand has an infrared LED, a night vision camera captured these movements, and streams the video to Raspberry Pi. The patterns are recognized by machine learning models, and using Raspberry Pi, the tasks are performed. Number four, meet Auto Bill, a fast and effective setup to generate shopping bills, thus reducing the checkout time. This project might look like a product, but it's completely DIY. It identifies the objects using a camera placed above, while the load cell at the bottom weighs the object. The object along with its weight, price, and quantity is then automatically added to the cart, and the bills generated, thus emitting the need for human involvement. The users can simply pay the total amount and collect their groceries.

thus, it reduces the chances of forming long cues. Number three, if you're bored being alone at home, then this Nerf turret bot can become your friend. But watch out as it'll throw turrets at your face. The turret continuously tracks the face, and the Python script sends the face coordinates to the Arduino. The Arduino then moves a Servo in order to position the camera at the center of the face. As the face is detected, turrets are thrown. All this process happens so fast that the turret barely misses a target. So, would you dare to try this project out? Tell us in the comments. Number two, one of the common causes of pollution is improper waste management at the starting level, for example, homes. We often mix various types of waste in one bin, which gets difficult to segregate at a higher level. That's where this project can help. This smart Dustbin made by Samuel identifies the type of waste thrown at it and then accordingly rotates the bottom section of the bin to collect that, whether it be paper, plastic, or any other type of waste. The smart bin takes care

of all things number one. This is another project to improve your high score in certain games. The setup might look simple, but can get the job done for you. At the top, it has a camera placed inside a frame that continuously looks at the screen, while the bottom section has an arm mechanism connected to the controller's joystick. Using machine learning, the bot's able to identify the object on the screen. The Arduino then moves the joystick over the object using a Servo motor and plays the games easily. So, these were some of the best machine learning projects that you can try. Drop a like and subscribe to our channel to keep supporting us. We'll be back with some great project ideas soon. Till then, goodbye.

12 New AI Projects using Raspberry-Pi, Jetson Nano & more

ToP Projects Compilation

혁신적인 머신 러닝 프로젝트 아이디어

PID 컨트롤러를 통해 몸의 움직임을 이용해 드론을 제어합니다.

안경에 카메라를 장착하여 2계층 신경망으로 제스처를 통해 가전제품을 제어하기.

전방향 휠이 있는 물리적 설정을 사용하여 픽셀 스캐닝 에임봇을 만들어 게임 정확도를 향상시킵니다.

카메라 및 모션 기술을 활용한 혁신적인 프로젝트

- 메탈 기어 모터와 인텔 리얼센스 카메라를 사용하는 에임봇은 양궁에서 과녁의 색상과 깊이를 계산하여 과녁을 쉽게 맞출 수 있습니다.

- 강력한 PCB 설계 소프트웨어인 Altium Designer는 이제 Octopart를 통해 Altium 365와 설계를 공유하고 실시간 부품 인사이트에 액세스할 수 있는 기능을 제공합니다.

- 신체의 움직임으로 드론을 제어하는 프로젝트는 안전 기능을 강화하여 특정 조건이 충족될 때만 작동하도록 했습니다.

기술 애호가를 위한 혁신적인 DIY 프로젝트

적외선 LED와 라즈베리 파이의 머신 러닝을 사용하여 공중에 그려진 글자를 기반으로 작업을 수행하는 지팡이.

물건과 무게를 식별하여 쇼핑 계산서를 생성하여 계산 시간을 줄여주는 DIY 설정인 AutoBill.

파이썬 스크립트와 아두이노로 제어하여 얼굴을 추적하고 터렛을 던지는 너프 터렛 봇은 독특한 인터랙티브 경험을 제공합니다.

DIY 게임 봇: 머신러닝과 아두이노 사용

DIY 게임 봇은 카메라와 머신러닝을 사용하여 화면의 물체를 식별하고 서보 모터를 사용하여 조이스틱을 움직입니다.

이를 통해 봇이 효과적으로 게임을 플레이할 수 있으며 재미있는 머신 러닝 프로젝트가 될 수 있습니다.

더 많은 프로젝트 아이디어와 튜토리얼을 구독하세요.