搜尋結果

AI Self-Driving Car Racing Competition Attracts High School Students from North, Central, and South In collaboration with the brand AI4kids, a summer event was held at the Robotics Maker Base in the Central Taiwan Science Park. Utilizing NVIDIA Jetson Nano edge computing, a four-day AI self-driving car practical competition was designed, drawing participants from Taipei, New Taipei, Hsinchu, Taichung, Yunlin, Kaohsiung, and Penghu. Guided by professional instructors and the AI4kids teaching team, students progressed from foundational machine learning and deep learning knowledge to hands-on experience with JetBot car hardware assembly, motion control (basic programming, remote control), intelligent control (obstacle avoidance, object tracking), and finally, independent training of AI visual recognition models. The event concluded with a Jetbot self-driving car race! How to Equip the Racing Car with AI Vision? The self-driving car race track is a square circuit, and the challenge for the students is to enable the car to complete a lap in the shortest time possible using computer vision without any manual remote control or programming. In the process, teams of two first had to complete track image collection, AI model training, and refinement. They underwent multiple testing and verification phases with parameter adjustments. Many students were trying to train their own AI model for the first time, encountering challenges such as reflections on the track during image capture, uneven road surfaces, and parameter setting errors leading to unstable car control. Multiple adjustments were necessary to successfully make the car complete a lap, and students who spent half a day on training exclaimed, "Time is tight!" "We need to repeatedly collect samples, gather images from various angles, and plan for various movements. We took more than 200 pictures for learning and repeated the training 50 times. During training, we utilized the built-in GPU of Jetson. In the end, we achieved the car's movement. The final presentation involves four closely connected data points: 'overall speed,' 'turning acceleration,' 'track trajectory differentiation processing,' and 'balance of left and right motors.' Integrating 'environment,' 'venue,' 'data,' and 'four major variables,' completing a simple track is really not easy!" — Chen, a student from Zhengxin High School Competition Champion - "Realized It Wasn't as Easy as Imagined!" "Our team's Jetbot encountered some problems, so we started testing a bit later. We faced many issues during the production, such as being unable to connect to the computer or finding the robot on the computer. In the afternoon, while operating the robot and capturing photos for training the model, we encountered consecutive issues where the program couldn't proceed or errors occurred. Additionally, the robot unexpectedly ran out of power. The only solution was for us to slowly identify the causes of the problems and troubleshoot the errors. We also had to find the corresponding charger to recharge the robot." — Wu, a student from Xinmin High School Jetbot Self-driving Car Video Recordings Not only practical courses but also comprehensive foundational knowledge! On the first day of the four-day camp, participants were introduced to the field of artificial intelligence, covering topics such as neural networks, machine learning, the Internet of Things, and the Python programming language. The second day advanced the understanding of the robotic world, including embedded single-board computers (Jetson Family), robot brains, and visual installations (OS & Camera), deep learning (NVIDIA Jetson-inference), and computer vision applications (detection, recognition, segmentation). "On the second day, we used ThinkerCAD to simulate Arduino. In the afternoon, we introduced edge devices and the application and architecture of Jetson in daily life. Next, we experienced a robotic arm competition, where we used a robotic arm to pick up yellow rubber ducks, and the one who picked up the most was the winner. After that, we assembled the JetBot together. Fortunately, the distribution inside wasn't too difficult, and I could understand it by reviewing a bit. During the course, I learned about some Arduino components and gained theoretical knowledge. Although I had encountered some of the topics in school, it was still interesting to listen to." — Wei, a student from Mingdao High School Learning Portfolio, Providing Comprehensive Feedback on the Learning Process and Achievements! At the end of each day's sessions, the camp allocated ten minutes for participants to review the key points of the day's lessons, reflect on the absorbed knowledge, and write a learning portfolio, creating a comprehensive record of their learning process. In addition, the camp provided hands-on guidance to help each participant obtain the international certification "Getting Started with AI on Jetson" from NVIDIA, enriching their four-day learning journey! Four-Day Camp, Broadening the Horizon of High School Students in AI! The four days of courses truly brought many different experiences and intellectual stimulations to each student. Opening the door to the field of AI robotics and autonomous vehicles for the participants, the camp also hopes that each student will explore more deeply in the future and create more applications related to autonomous vehicles! "Four days ago, I had never encountered robots or Jetbot, but with the help of the teaching assistants and teachers, I successfully completed the tasks. It was very fulfilling, and I learned a lot of knowledge that I didn't understand before. I also had the hands-on experience of assembling Jetbot, which was unprecedented for me. If I have the opportunity to work on this again in the future, it should spark my interest." ——Nanhu High School, Hong Student

If AI is important, where should children start, and how can parents inspire their interest in learning? Do they need a foundation in programming? Should children be introduced to various AI algorithms? Chen Sheng-Wei, Executive Director of the Taiwan Artificial Intelligence School and Chief Technology Officer of Yuanta Financial Holdings, offers advice to parents: "The technology of AI makes a major breakthrough every three months. By the time children graduate from university and enter society, AI will be significantly different from today. Therefore, there's no need for rote teaching or in-depth mathematical theories. The first priority is to make children unafraid of computers and create a well-designed learning environment with situational teaching. Let children naturally understand how AI works and feel how AI will connect with their lives in the future." "For routine tasks in life, encourage children to think and use AI. Through simple programming tools like Scratch and Arduino, let them assemble and design solutions on their own. This is a way parents or teachers can design learning methods that involve children in AI." Related Videos:https://www.youtube.com/watch?v=DhY605-0hjk&t=30s Integrating the Ministry of Education's 108 Information Technology Curriculum, the U.S. CSTA K12 Computer Science Learning Framework, and recommendations from the American Association for Artificial Intelligence (AAAI), comprehensive AI learning from Grade 1 (K1) to Grade 12 (K12) should encompass six major processes: Computational Thinking According to the curriculum design by the Ministry of Education, Computational Thinking aims to cultivate children's ability to think using computational tools, enabling them to analyze problems, develop problem-solving methods, and make effective decisions. Programming is considered a feasible method for cultivating Computational Thinking. Perception Introduce children to sensor components, help them understand how computers perceive the world through sensors, and explore various sensors such as visual, auditory, vibration, light, temperature, etc. Understand the differences and limitations between computer perception and human perception. Representation and Reasoning Teach knowledge representation and various reasoning algorithms. Explore how computers classify and recognize information, leading to inference and decision-making. Examples include simple classification, multiple classifications, and the application of decision tree algorithms. Machine Learning Machine Learning is a core technology of Artificial Intelligence. Through practical implementation, children can learn that machines can learn from data. Explore supervised learning, unsupervised learning, and reinforcement learning as three machine learning methods. Understand the basics of neural network architecture and how the choice of training data affects machine learning outcomes. Natural Interaction Explore the integration of multiple artificial intelligence technologies, such as image recognition, speech recognition, and automated robots, to achieve natural interaction with humans. Allow children to interact with intelligent devices to understand the principles behind the operation of robots and be aware of their limitations. Social Impact Encourage children to pay attention to biases and ethical issues related to artificial intelligence. Discuss potential positive and negative impacts on society. As future citizens, they should maintain critical thinking skills and make informed choices regarding the development and use of AI technologies in social construction and services. In the next decade, AI technology will accelerate its integration into various industries and become an integral part of our daily lives. The six learning processes in AI education for K-12 students aim to empower them with critical thinking and creativity. With practical implementation, children can explore opportunities in various fields where AI technology will play a crucial role. AI will create new opportunities, and technological advancements will delegate repetitive and complex tasks to technology. This will give individuals more time to focus on meaningful activities. Considering alternative methods of professional evaluation and the time invested daily, individuals should reflect on their priorities and passions to find a meaningful direction for their future.

Implementing a Green Lizard Detection and Prevention Dashboard 1. Project Overview This page serves as an introduction to the open-source project "Green Lizard Detection and Monitoring" within the Nvidia Jetson community. The project utilizes Jetson Nano in conjunction with computer vision-based object detection models to real-time track and present the movements of green lizards. We also provide the Github repository for this open-source project, allowing readers to replicate the project for quick development when working on similar projects. 2. Project Motivation What is the Green Lizard? The Green Lizard, also known as the American Iguana, is a large lizard that lives in trees. It can reach a total length of 1 to 2 meters from head to tail and has a lifespan of over 10 years. The lizard primarily feeds on plant leaves, shoots, flowers, and fruits. It is a diurnal reptile and can lay 24–45 eggs per clutch. Known for its strong reproductive ability and adaptability to different environments, the green lizard poses ecological challenges. Significance of the Green Lizard Ecological Issue In Taiwan, abandoned green lizards rapidly reproduce in the wild. The Forestry Bureau of the Council of Agriculture in Taiwan has expressed concerns over the ecological impact of green lizards. They not only harm the ecosystem but also cause agricultural losses by consuming crops. Additionally, their habit of digging holes, especially near riverbanks and fish ponds, damages infrastructure. According to the government's current monitoring efforts, the presence of green lizards has spread to multiple counties and cities. Establishment of a Green Lizard Monitoring System With the recent advancements in edge devices, we have the opportunity to develop and deploy real-time computer vision applications for detecting and monitoring the movements of green lizards. Through the implementation of a system with real-time alerts and notifications, we can take timely measures to prevent asset losses caused by green lizards. Software Architecture 3. Data Collection Green Lizard Image Labeling Illustration Image Crawler Through a dynamic web crawler using Selenium, we collected around 5000 images of green lizards and individually labeled each one. We also provide downloads for the image and label data; please refer to our image labeling tutorial. Image Labeling For collaborative labeling, we used the web-based object detection labeling tool makesense.ai due to its simple and user-friendly features, eliminating the need for any software installation. You can find the tutorial documentation for using this tool at this link. 4. Training the Green Iguana Detection Model using Nvidia TLT Nvidia Transfer Learning Toolkit Introduction to Nvidia Transfer Learning Toolkit (TLT) Creating an AI/ML model from scratch to solve business problems is a highly costly process. Therefore, transfer learning is often used during project development to expedite the process. Transfer learning is a subfield of machine learning that focuses on leveraging existing solution models, such as pre-trained models, for different but related problems. NVIDIA's Transfer Learning Toolkit (TLT) is precisely a toolkit for transfer learning, providing popular pre-trained models for image and natural language processing tasks. TLT is primarily designed for deployment purposes, integrating features for size reduction and optimization of models after training. In the ideal scenario, using TLT for model training only requires setting training parameters, eliminating the need to write code from scratch. TLT realizes an end-to-end process for training, optimization, and deployment, significantly differing from typical development workflows and greatly reducing project development time and technical costs. Hardware Requirements The recommended hardware configuration to run Nvidia Transfer Learning Toolkit (TLT) is as follows: - 32 GB system RAM - 32 GB of GPU RAM - 8-core CPU - 1 NVIDIA GPU - 100 GB of SSD space Software Requirements NVIDIA TLT Installation Guide We provide a detailed installation guide for NVIDIA TLT. Please refer to our prepared environment setup tutorial for step-by-step instructions. TLT Model Training and Optimization The object detection model used in this project is YOLOv4. You can refer to the YOLOv4 training example code with slight modifications to train the model using the images and labels of green iguanas. After running the code and waiting for the completion of the process, it will automatically export a .etlt file, which is the final model file for deployment. Note: Sample code for quick hands-on experience with TLT's computer vision functionalities can be found in the official link. Most of the examples require minimal parameter adjustments for one-click training, optimization, and export. 5. Deploying the Model on Jetson Devices Introduction to Deepstream After completing model training, the next step is to deploy it on Jetson devices for real-time inference, while also streaming images and inference results. We use Nvidia DeepStream, a framework specifically designed for inference and data streaming. DeepStream is an end-to-end framework that facilitates deep learning inference, image and sensor processing, and the transmission of insights to the cloud in streaming applications. You can construct cloud-native DeepStream applications using containers and deploy them at scale through Kubernetes coordination. When deployed at the edge, applications can communicate between IoT devices and cloud-standard message brokers (such as Kafka and MQTT) for extensive wide-area deployments. Introduction to TensorRT NVIDIA TensorRT™ is an SDK designed for deep learning inference. The SDK includes a deep learning optimizer runtime environment, enabling deep learning applications to achieve low latency and high throughput. Before deploying a model, TensorRT is used to transform the model, accelerating the runtime speed during inference. TLT, oriented towards deployment, is a model training framework that inherently includes transformation capabilities for post-training models. Simply using the tlt-converter allows the TensorRT inference engine to build from the .etlt output model trained by TLT. Additionally, hardware and software for deployment on devices require conversion using the corresponding version of tlt-converter. The following are all configuration options: Installing DeepStream Environment We have documented the steps for installing the DeepStream environment, as well as the process of building the TensorRT engine for the .etlt model. This includes methods for installing other packages needed for this project, such as MQTT client, and more. For detailed installation steps, please refer to our prepared environment setup tutorial. Running DeepStream deepstream_app/deepstream_mqtt_rtsp_out.py is the configuration for running deepstream python binding. This program performs object detection inference, streams images in RTSP format, and sends the resulting inference data to the MQTT broker, storing it in the database for future visualization. 6. Data Dashboard Plotly Dash Plotly Dash is a framework for interactive data visualization. It allows writing a frontend page for data visualization using non-front-end languages. The other sub-packages extended by this package simplify the web style and layout process, making it ideal for quickly developing web-based data visualization. In addition, the live-update feature provided by this package allows the web page to continuously update dynamically, making it especially suitable for the dashboard requirements of this project. Dash Bootstrap Components Bootstrap is an open-source front-end framework for website and web application development. It includes frameworks for HTML, CSS, and JavaScript, providing fonts, forms, buttons, navigation, and various components, as well as JavaScript extension packages. Its goal is to make the development of dynamic web pages and web applications easier. Using Dash Bootstrap Components allows for the direct inheritance of the frameworks defined by Bootstrap, enabling the quick and easy layout configuration of web pages. MQTT broker Eclipse Mosquitto Users can choose an MQTT broker based on their preferences. In this project, we use Eclipse Mosquitto. Once installed, simply start the MQTT broker to begin receiving data. Database Setup MySQL This project uses MySQL to set up the database. Users can choose their preferred database type. The purpose of the database is to write the data flow from Jetson Nano to the MQTT broker into the database. Plotly Dash periodically updates the data and presents the latest data on the dashboard. MQTT Topic Subscription and Data Writing to Database When Deepstream is running on Jetson Nano, it continuously sends inference result data to the MQTT broker through the MQTT communication protocol. By subscribing to this specific topic, you can receive notifications when the latest data flows into a particular topic and write the data to the database. The script for this process is located at mqtt_topic_subscribe/mqtt_msg_to_db.py. Run the Web Dashboard Deepstream Deployment and Execution MQTT broker runs on the data server Database runs on the data server With the above conditions, the web dashboard can start running. Please note that the configuration of Deepstream is adjustable, so if users want to replace the model with another object detection model in the future, they only need to modify the configuration files and Deepstream Python executable.

Introduction Badminton is a sport characterized by high levels of skill and speed, requiring players to possess exceptional technical abilities and outstanding reaction times. However, beyond individual skills and responsiveness, a player's tactics and strategy also play a crucial role in determining the outcome of a match. In order to elevate the overall athletic level and performance of players, badminton coaches have been continuously seeking new ways to analyze match data and player performances. Today, with the constant development and innovation in technology, badminton big data and computer vision technology have gradually become the secret weapons of badminton coaches. In this article, we will share the applications and advantages of badminton big data and computer vision technology in the sport of badminton. We'll introduce the fundamental principles and practical uses of these technologies, as well as their impact on player training and match performance. Computer Vision and Image Processing Image processing is a technique used to enhance, improve, or transform images. In badminton, image processing can be applied to frame analysis, trajectory visualization, video frame restoration, and video processing and editing. Frame analysis involves breaking down a video into individual static frames and extracting valuable information such as player positions, shuttlecock trajectories, and hitting techniques. Trajectory visualization allows for the visual representation of this information, enabling players and coaches to better analyze the problems and strengths during matches and training sessions. Video frame restoration can enhance incomplete or blurry images, providing a clearer view of the game and training situations. Video processing and editing allow the compilation of multiple video segments into a comprehensive video, showcasing the highlights of matches and training sessions. These technologies can assist players and coaches in better understanding the challenges and advantages during matches and training sessions, ultimately enhancing the players' competitive levels and overall sports experience. Badminton Big Data and Tactical Analysis Badminton big data refers to a large amount of data collected from badminton matches and training sessions. This data includes players' personal information, match results, technical indicators, etc., helping players and coaches better understand player performance and training effectiveness. Through tactical analysis and visualization, we can intuitively understand a player's strengths and weaknesses, thereby developing corresponding tactics and training plans. Practicing code for the collection and analysis of technical and tactical data can assist individuals in mastering these technologies more effectively. Tactical analysis and visualization involve statistically analyzing and visually presenting data collected during badminton matches and training sessions. Through analysis and visualization, players and coaches can gain a clearer understanding of a player's strengths and weaknesses, such as scoring rates, error rates, and successful attack rates. This data can aid in formulating effective training plans to further enhance a player's competitive level. In addition, practicing code for the collection and analysis of technical and tactical data involves utilizing programming to implement the collection and analysis of badminton big data. By learning programming, individuals can better grasp the technical aspects of badminton big data and apply them to player training and competitions. In the field of badminton, TrackNet is a commonly used image processing technology that captures and analyzes the trajectory of the shuttlecock in images. This technology helps players and coaches better understand the situations during matches and training sessions, ultimately improving the level of play. Conclusion In this article, we introduced the applications and advantages of badminton big data and computer vision technology in the sport of badminton, as well as their impact on player training and match performance. These technologies not only help players analyze their own performance and identify areas for improvement but also assist coaches in more effectively analyzing match data and devising tactical strategies. Through badminton big data and computer vision technology, badminton coaches can analyze players' performances and match data more comprehensively and accurately, leading to the formulation of more effective training plans and tactical strategies. Simultaneously, these technologies help players better understand their performances and progress, enabling them to adjust their training and match strategies more effectively. In conclusion, badminton big data and computer vision technology serve as secret weapons for modern badminton coaches and essential tools for improving overall player competitiveness and match performance. We believe that with ongoing technological development and innovation, these technologies will play an increasingly important role in the future of the sport of badminton. Interested in learning more? AI Sports Applications Online Course: https://ai4kids.ai/courses/ai-sport/ In collaboration with the brand AI4kids and Professor Yi-Wei Yeh's team at National Yang Ming Chiao Tung University, this online course provides video tutorials to step into the interdisciplinary field of AI and sports through systematic learning. AI Sports Specialized Camp: https://ai4kids.ai/ai-camp-ai-for-sport/ Led by professors from National Yang Ming Chiao Tung University and National Chung Hsing University, along with AI experts, this 4-day camp allows you to delve into the fundamental concepts and technologies of machine learning and deep learning. You will gain hands-on experience in the latest applications of AI in sports research.

AI is about to change the world, but who will change AI? — Stanford University Professor Fei-Fei Li According to AI4k12.org, an organization formed by the American Association for the Advancement of Artificial Intelligence (AAAI) and the Computer Science Teachers Association (CSTA), there are four compelling reasons presented in 2019 advocating for K-12 students (from 1st grade to 12th grade) to learn AI (Artificial Intelligence) now: Reason 1: AI plays an increasingly crucial role in today's society, including smartphone voice assistants, autonomous vehicles, and intelligent robots in the workplace (sometimes at home). Reason 2: Future citizens need to understand the basic knowledge of how our society applies AI and how essential public policies related to AI technology are made. Reason 3: AI technology will lead to unemployment in certain fields while creating new employment opportunities in others. Reason 4: Due to the growing demand for workers with AI knowledge in society, students should be encouraged to learn in a STEAM (Science, Technology, Engineering, Arts, and Mathematics) context from an early age. In response to the AI trend, the Ministry of Education has also added the "Curriculum Guidelines for Technology Education" in the 12-Year Basic Education Program. In the new 108 curriculum, "Programming" is listed as a required course for both junior and senior high schools. The aim is to cultivate computational thinking, logical thinking, and systematic thinking through the learning of computer science-related skills.

Since the implementation of the 108 Curriculum Guidelines, in addition to the subjects tested in the college entrance exams, there is now the inclusion of "learning process" as an indicator for university applications. This allows students to assess their personal inclinations and future directions through their learning experiences, serving as a basis for choosing future academic paths. The Ministry of Education promotes the learning process because it aims to encourage students not only to focus on academic studies but also to participate in extracurricular activities such as external competitions, volunteering, obtaining certifications, academic learning, summer camps, and winter camps. As the holidays approach, we have compiled a list of AI-themed camp activities offered by prestigious universities, including camps at National Taiwan University, National Tsing Hua University, National Chiao Tung University, and more. Additionally, we have included online Python learning courses. These opportunities allow junior and senior high school students to not only experience practical applications in various academic fields during the holidays but also to make new friends and enrich their holiday experiences through these activities. 2023 High School AI Programming Camp - Early Bird Discount There are countless camp programs in Taiwan, catering to children with different interests and inclinations. Here, we recommend a variety of AI-themed camps as a quick reference for those interested in artificial intelligence. 2023 Winter AI Medical Project Implementation Camp Taught by the MeDA Lab research team from National Taiwan University (led by Professor Wang Weizhong), physicians, and AI experts, this camp delves into the analysis and application of medical imaging and big data. Participants gain insights into how big data and AI will impact the future of healthcare in Taiwan. The program includes expert-led sessions and a hackathon competition where participants can showcase their creativity, logical thinking, and collaborative teamwork to create their own AI healthcare projects. Camp Focus: Fundamental knowledge of artificial intelligence Practical applications of AI in medicine AI model training Programming implementation AI healthcare project competition Suitable for ages: 15-19 (Junior high school third grade, high school, and college freshmen) Supported by the Precision Sports Program of the Ministry of Science and Technology, Professor Yi Chih-Wei's team from the Department of Computer Science at Yang Ming Chiao Tung University will lead participants in exploring the practical applications of AI in sports events. This includes data collection systems, simulating the skills of top players with an intelligent serving machine, and analyzing player positioning and landing points. During the course, participants will have the opportunity to experience the operation of AI serving machines, real-time sports body recognition, player positioning analysis, and shuttlecock landing point analysis on the badminton court. The program concludes with group collaboration to collectively complete a sports project. Camp Focus: Fundamental knowledge of artificial intelligence AI model training implementation AI badminton court experience AI sports project competition Led by Professor Su Li from Academia Sinica's research team, Professor Lin Yijun, the director of the AI Orchestra at National Tsing Hua University, and the Forgetful Technology team, participants will enter the world of AI music. They will gain insights into technologies such as AI automatic scoring, automatic accompaniment, and virtual musicianship. The program includes the appreciation of chamber music performed in collaboration with AI. In the final stages of the course, experts will guide participants in operating Vtuber creation tools and AI music compilation, enabling them to create their own Vtuber virtual personas for performances. Basic knowledge of artificial intelligence Implementation of automatic scoring and chord generation Experience in chamber music ensemble Creation of Vtuber and personalized music Python Online Course: $1,200 - No time limit Python Data Science Course: $2,000 - No time limit Machine Learning Course: $2,000 - No time limit AI Visual Recognition Course: $2,000 - No time limit 1. Record Your Learning Journey Participating in camps or online courses allows you to document your learning journey, and this is not just about earning extra points. Therefore, it's not about joining a multitude of camps; instead, it is recommended to focus on camps related to your future development. For example, if you aspire to become a doctor and aim to enroll in a medical school, having participated in a medical camp at a prestigious university during high school, and being able to express your motivations and experiences coherently, is convincing! 2. Clarify Your Interest in the Field and Identify Your Life Direction As mentioned earlier, brief camp experiences can help confirm whether you truly enjoy a particular field. If you determine that it's not the direction you want to pursue, that's okay; at least you've avoided going down the wrong path. Through diverse experiences, you can eventually find a direction that you both enjoy and find suitable! 3. Enhance Interpersonal Relationships, Foster Teamwork, and Leadership Skills Participating in camps not only provides academic knowledge and hands-on experience but also teaches interpersonal skills, cultivates leadership abilities, and promotes teamwork. Additionally, camps bring together peers from all over Taiwan, allowing you to meet elite friends from different regions. After participating in various camps, you can sense the unique personalities of individuals and learn how to interact with people of different temperaments. Moreover, academic exchanges during the camps facilitate mutual learning! Conclusion The recommendations above include diverse AI artificial intelligence camps and additional suggestions for online Python learning courses. It is emphasized how participating in camps can benefit you by developing your skills and helping you establish your future direction. Hopefully, everyone can make the most of their holiday by exploring the camps and online courses that interest them. If you come across a camp or online course that piques your interest, feel free to click on the relevant information for registration!