In the medical imaging AI industry, an increasing number of startups are emerging, among which AetherAI has recently attracted attention At last year's MICCAI, an international medical imaging conference, AetherAI's digital pathology AI defeated Stanford University's team, aiming to utilize artificial intelligence to achieve precision medicine and alleviate the burden of time-consuming tasks on physicians Providing digital pathology solutions to meet artificial intelligence application needs Led by Dr Zhao-Yuan Yeh, AetherAI, though only a few years old, includes members skilled in both healthcare and technology, possessing strong interdisciplinary integration capabilities They excel in medical research, data science, software development, systems engineering, and medical knowledge and information technology, committed to offering solutions for digital transformation in pathology and AI-assisted diagnostics With digital pathology becoming a milestone in the development of whole-slide imaging technology, AetherAI has introduced the aetherSlide system solution Besides developing digital slide management and viewing systems, it also integrates image annotation, deep neural network inference, and training functionalities, thereby fulfilling the needs for AI module applications and development Although AetherAI has been formed only a few years ago, it has recently gained significant attention in the medical imaging AI industry Just recently, Dr Zhao-Yuan Yeh, Co-founder and CEO, was awarded as one of Taiwan's Top 100 MVP Managers Photo credit AetherAI's Facebook page A key feature of the digital pathology system is the customizable digital slide status bar, which can sort cases by priority, urgency, etc, thus providing clear visibility and facilitating time management The interface is also user-friendly, offering hotkeys combinations, one-click slide assembly, along with tools such as rulers, magnifiers, and seamless rotations that enhance consultation or discussion efficiency Regarding AI services, it offers applications like cancer detection and quantification, immunostaining quantification, and blood cell classification counting Cooperation between AI and physicians minimizes repetitive work, and the system supports extensive training annotation features with various selection modes, multi-category tagging, and freehand drawing Annotations integrate seamlessly into deep learning training, structured format output, and AI training data generation in everyday processes Last year, AetherAI developed an AI medical imaging development platform aetherAI, notable for its diagnostic automation, providing an end-to-end digital pathology AI development process Photo credit AetherAI's Facebook page It is worth mentioning that the digital pathology system supports robust management functions, especially capable of being integrated according to the hospital department's work assignment process, and even with existing hospital information systems, this saves man-power, enhances administrative efficiency through digitization In terms of file formats, it supports multiple brands of slide scanner types and whole-slide image formats like svs, ndpi, scn, mrxs, bif, tif Additionally, to reduce the burden of long-term storage in medical facilities, AetherAI's infrastructure supports significant expansion capabilities, offering scalable options based on user needs, and supporting local and data center options for long-term storage solutions AetherAI's digital pathology system boasts strong management features and supports various brands of slide scanners, enhancing workflow efficiency through digital operations AetherAI Digital Pathology AI Applications Reduce Doctor's Burden and Enhance Productivity and Consistency At the recent AI HUB conference, AetherAI demonstrated its AI medical imaging development platform launched last year aetherAI, its main feature being diagnostic automation This allows departments within hospitals to integrate various types of DICOM files and medical knowledge, boasting a highly scalable AI model capability, and providing an end-to-end digital pathology AI development workflow Currently, it offers digital pathology AI modules including automated bone marrow smear classification, nasopharyngeal carcinoma recognition, and glomerulus detection applications, involving nearly ten different types of datasets So, what are the tangible benefits for doctors Simply put, with a prior scan using AI, cancers can be confirmed without the lengthy manual review previously typical in bone marrow exams, thus greatly shortening repetitive tasks for doctors and enhancing efficiency in complex diagnoses Currently, aetherAI has reached recognition levels comparable to a pathology doctor's visual assessment standards, achieving an identification rate as high as 97 for nasopharyngeal cancer AetherAI's AI medical imaging development platform aetherAI can significantly reduce repetitive tasks for doctors, leading to more efficient and effective high-complexity diagnoses Currently, AetherAI's partners and customers mainly include large medical centers, with the University of Pittsburgh Medical Center leading internationally In Taiwan, includes major medical institutions such as Taipei National University Hospital, Taipei Veterans General Hospital, Chang Gung Hospital, Cathay General Hospital, Tri-Service General Hospital, Chung Shan Medical University Hospital, Taipei Medical University Hospital, among others, aiming to use artificial intelligence for precise medical applications, enabling deep learning in clinical practice to reduce the workload for doctors and elevate the consistency of medical quality AetherAI official website「Translated content is generated by ChatGPT and is for reference only Translation date：2024-05-19」

In today's digital age, both individuals and companies face challenges in writing copy and designing layouts when managing a brand online What constitutes good design The online world often gives a mysterious impression, and the WoodThought team has been focusing on studying various data sources for years, hoping to find the answers they seek Focused on data analysis for exploring and solving problems WoodThought is a group of consultants specializing in data Big Data amp Deep Learning and solving the essence of problems The team utilizes various quantitative analysis techniques to challenge many of the profound issues found on the internet and strives to continually enhance the data analysis capabilities of both Taiwanese businesses and individuals The application of this technology is extensive, including personalized navigation and recommendation systems on news platforms, eCommerce personalization for navigation, search, and recommendations, or strategy development, design, and backtesting in financial transactions Moreover, WoodThought also offers data analysis courses like web scraping and data visualization in response to technological and current events, aiming to cooperate with both domestic and international enterprises to implement various data science solutions WoodThought is a group of consultants dedicated to data analysis, exploring and solving fundamental issues, utilizing various quantitative analysis techniques to challenge the many uncertainties of the internetImage Source） Taking the simplest web design as an example, WoodThought believes that with the advancement of personalized tracking technology and the combination of AB Testing experimental designs and testing techniques, professionals in all industries can now avoid relying on the so-called '20 years of marketing experience' Instead of using inefficient and unpredictable methods, making use of data analysis enables even intuitive observations to have a basis, thereby capturing the sentiments of online users Supported by adequate data evidence, incorrect decisions and directions can be corrected promptly before mistakes are made WoodThought showcased its proprietary 3D marking system at the recent AI HUB conference, which utilizes AI image recognition technology to help doctors quickly determine the condition of patients' lung nodules The AI model will automatically learn from the markings, providing recommendations for future markings by doctors, allowing for early discovery and immediate treatment, resolving past difficulties of 3D image marking and the challenge of obtaining marking data The team developed a 3D marking system using AI image recognition technology, significantly enhancing the efficiency of medical diagnosis The same AI data analysis technology, also applicable in the medical sector, was showcased by WoodThought at the recent AI HUB conference This includes their own 3D marking system with features like Auto-Learning and Pre-Labeling This system assists doctors in diagnosing lung nodules As doctors complete the diagnosis and marking, the AI model will learn from it and provide future marking suggestions, enabling early discovery and immediate treatment, thereby greatly resolving the difficulties of 3D imaging marking and issues of data accessibility WoodThought aims to solve various personalized service demands through data analysis techniques and calls for everyone to go beyond mere imagination, interact personally with different types of data, and dig out the answers behind the problems Over the past two years, core members of the WoodThought team have actively assisted and guided ordinary users on various e-commerce and media platforms to establish correct views through introducing relevant technologies and services This helps solve various personalized service demand issues and enhances the likelihood of matching products with customers WoodThought encourages everyone to leave behind imaginary notions, to engage with data, and to uncover the answers behind the problems, while also experiencing the problems behind the answers Given the rapid changes of the internet era, each rising wave brings different user needs and voices WoodThought aims to help everyone solve problems quickly and efficiently using various data analysis techniques「Translated content is generated by ChatGPT and is for reference only Translation date：2024-05-19」

Inspection is an important part of maintaining quality in all factory production lines However, manual inspection is traditionally used to inspect the appearance of products for defects Visual inspection is not only prone to errors, but also requires high labor costs Spingence implemented AI technology into the inspection system and developed a high-speed AI defect inspection system for passive components, hoping to greatly improve the speed and efficiency of the inspection process Focusing on automation software development to improve production line inspection efficiency Spingence focuses on software design, development and sales It has a comprehensive and easy-to-understand automation development platform It uses the automation graphical platform LabVIEW and high-end software to provide evaluation planning for automation projects, including robotic arms, machine vision, and motion control equipment, in hopes of changing the existing automation software model through various implementation cases and experience It provides a production line inspection platform that can be quickly introduced and is easy to use, which helps major companies improve the automation process of their production line Spingence focuses on software design, development and sales, provides a production line inspection platform that can be quickly introduced, and assists major companies in improving the production line automation process Source of image Spingence At the recent AI HUB conference, Spingence displayed its high-speed AI defect inspection system for passive components As the name implies, the inspection system introduces AI technology to make the entire production line more efficient The reason for developing the equipment is because most factories used to use manual inspection of the appearance of products for defects, but long hours of manual inspection will result in a decline in work quality In addition, as more components become smaller and production speeds increase, it will become unbearable to the human eye This is why we hope to use the power of technology and use high-resolution cameras and high-performance imaging software to make inspection work of the entire process more efficient Rapid recognition and self-learning meet the high-speed inspection requirements of production lines The high-speed AI defect detection system for passive components can easily complete automatic process editing and arrangement based on different hardware configurations, and also provides optimized models Its speed can reach 1,200 pieces per minute with missed detection rate below 50ppm More importantly is the flexibility of configuration, in which different edge devices can be selected according to customersrsquo cost and speed requirements Of course, there is no problem in mixing them Taking the defect detection demonstrated on site as an example, the integrity of glue dispensed on the white platform is detected If traditional AOI algorithm is used, the surface of the glue may reflect light due to ambient light, leading to many misjudgments However, if the automatic learning characteristics of AI neural network are used, then it will be able to stably differentiate between reflection and defects from incomplete dispensing The high-speed AI defect detection system for passive components can easily complete automatic process editing and arrangement according to different hardware configurations Highly flexible configurations can be built according to customers' cost and speed requirements To build a network model with high accuracy, it not only requires long-term training, but also optimizes parameters for different products Spingence has highly optimized the algorithm in embedded devices, allowing the high-speed AI defect detection system for passive components to inspect thousands of objects in just one minute, meeting the high-speed detection requirements of production lines Image recognition has become the most important application of AI, especially in defect detection on production lines AI has fast recognition and self-learning functions, which can more significantly improve overall implementation benefits The software platform and technology of Spingence have been certified by many partners, and the introduction of automation into factories has become a trend in recent years Due to the differences in the design of each production line, automation must be highly customized work As more resources are invested in the future, Spingence also hopes that this automation software platform will provide customers with more diverse solutions and help companies easily move towards Industry 40 Spingence's official website

Have you ever had an experience where, while randomly browsing online stores, it seems like the website can read your mind, constantly recommending products you want to buy Recommending the right product to the right customer at the right time For example, when you see a piece of clothing you like, similar items or accessories that could match it immediately appear below, and before you know it your shopping cart gets fuller and fuller The reason the webpage understands you so well is primarily due to the implementation of an AI engine that can automatically analyze customer preferences and recommend the right product at the right moment, invisibly increasing the consumer's purchasing intent and stabilizing the platform's customer base This is the expertise of Taimi Rosettaai Taimi Rosettaai provides software services to e-commerce platforms aimed at increasing conversion rates and customer retention rates, by analyzing consumer preferences through various AI engines and recommending personalized products online and offline Taimi Rosettaai is an internet company specialized in creating one-stop customer conversion and retention services for e-commerce, analyzing consumer behavior through various AI engines, and recommending personalized products both online and offline Referring to the previous mention of online shopping, throughout the entire shopping journey, clever integration of different AI engines on various situational pages—home page, product page, category page, shopping cart, checkout page—these designed APIs tailored to consumer needs and shopping contexts further deepen the analysis of consumer preferences Through accurate sales forecasting, it elevates the consumer experience and conversion rates Throughout the entire shopping journey, Taimi Rosettaai is capable of installing different AI engines on various pages, instantly recommending the right products in the right context, thereby invisibly enhancing consumer's purchasing intentions Different AI engines analyze different situations to create consumer return flow and increase lift-to-cart rates For instance, if a consumer lingers on a page buying a coat, not only similar styles get recommended below, but there may also appear clothing or accessories that complement that coat, such as undergarments, pants, belts, bags, etc The system will also consider the consumer's past shopping habits and recommend items they might be interested in For instance, if the consumer has previously purchased a DSLR camera, the page might suggest camera bags, cleansing kits, etc, relevant products Through AI technology, cross-analyzing merchandise information with customer preferences achieves personalized promotions Besides targeting the e-commerce market, Taimi Rosettaai also actively expands overseas, collaborating with major domestic and international accelerators in recent years, and hopes to apply its capabilities in different industries to seek more collaboration opportunities Taimi Rosettaai offers a variety of instantly usable API modules and basic situational combinations, allowing users to choose according to the characteristics of their website, without the need for large investments in building an AI infrastructure This enables clear understanding of customer preferences, significant reduction in marketing costs, and even inventory forecasting, thus effectively increasing e-commerce revenue Besides focusing on the e-commerce market, Taimi Rosettaai is also actively expanding overseas, having been selected by well-known domestic and international accelerators such as AppWorks, Founder Institute, Orange Fab, Zerothai, etc, over the past two years It also hopes to apply its capabilities to different industries in the future, seeking more collaborative opportunities Taimi Rosettaai official website「Translated content is generated by ChatGPT and is for reference only Translation date：2024-05-19」

For many professional team trainers, recording players' training conditions in the past often required manual effort However, by mounting sensors on sports equipment and combining them with corresponding AI devices, it's now possible to easily record training data Massive Digital Engineering has launched the napa smart badminton racket, digitizing the often abstract actions in sports, which assists athletes in finding the right training direction Introduction to Massive Digital Engineering Established in 2000, Massive Digital Engineering focuses on data analysis and mining in areas such as ERP, Industry 40, and financial big data It services manufacturing industries like chemicals, sports equipment, and automotive components, as well as various retail and distribution sectors by offering system development, customization, installation, and integration, helping companies enhance operational efficiency The company excels in cloud big data, using publicly traded financial databanks to design up to 92 financial assessment standards, which serve as benchmarks for businesses to improve their corporate structures Additionally, in response to Industry 40 and smart manufacturing trends, Massive Digital Engineering continues to develop innovative intelligent technologies aiming to boost production efficiency while reducing costs and pollution The napa smart badminton racket handle embeds high-performance sensors that automatically transmit the collected data for AI-driven analysis to an app Massive Digital Engineering showcased the napa smart badminton racket at the AI HUB Conference Recently at the AI HUB Conference, Massive Digital Engineering unveiled the extensively developed napa smart badminton racket, digitizing complex and hard-to-measure sports movements applicable in various competitive sports This tool helps coaches in training and assists athletes in finding the right direction For leisure and entertainment, the napa smart badminton racket can transform traditional courts into smart courts, complete with real-time data display boards on both sides of the court This not only displays scores during matches but also enables players to instantly access various swing data, vastly enhancing the fun and interactive aspect of the game With years of experience in data analysis and mining, one may wonder why Massive Digital Engineering ventured into the sports domain It turns out that there is a backstory involving long-term subcontracting work for the well-known sports brand WILSON, combined with the good reputation of its own Napa badminton rackets and familiarity with badminton-related products Hence, starting from the basic physical product of the racket, years of AI research were integrated into the development of the napa smart badminton racket project The sensors on the racket record all swinging actions including speed, posture, energy use, and striking force, and even 3D swing trajectories can be reviewed via the app But how does the napa smart badminton racket work Here's the principle sensors embedded in the racket handle automatically collect data, which, through AI algorithms, is linked to cloud big data When connected to a smartphone, various sports records can be viewed via an app The racket's sensors document every swing—its speed, energy consumption, posture, and power—and even 3D trajectories If used with a smart wristband on the same hand, it can also monitor heart rate and blood pressure, then through various big data applications, provide personalized scientific sports recommendations For instance, some players may swing too broadly with enough power but incorrect direction, or they may exert too much force during a swing but lack strength during impact These common training issues can be effectively addressed and improved through the napa smart badminton system Besides the napa smart badminton racket, the napa intelligent system can also be applied to baseball By installing sensors in the bat handle, swings can be recorded in real-time Smart sports application scenarios The primary target audience for napa smart badminton includes players and coaches, and it is also suitable for individuals who want to train independently Beyond badminton, the napa intelligent system can be applied to other sports like baseball The underlying principle is similar, hiding sensors in the bat handle to instantly record swing trajectories, enabling hitters to more precisely determine the impact position and power point Additionally, integrating more professional training, such as weightlifting—where Hsu Shu-ching won two Olympic gold medals—has gained more national attention Traditional training predominantly uses verbal instructions, such as directing athletes to apply a certain amount of force backward or forward, but these descriptions are abstract Integrating the napa intelligent system into weightlifting, for instance embedding miniature sensors in silicone gloves, would allow precise tracking of movement trajectories through app data, making adjustments more accurate with systematic data Massive Digital Engineering is actively collaborating with various badminton venues, aiming to upgrade traditional courts to smart courts using the napa smart badminton racket Massive Digital Engineering has recently started collaborating with various badminton venues to upgrade traditional courts to smart courts using the napa smart badminton racket They also continue promoting through sports communities, sports digitalization, experiential marketing, and other diverse applications to revolutionize the current sports industry「Translated content is generated by ChatGPT and is for reference only Translation date：2024-05-19」

As the population ages, home care has become a universal issue In the face of low medical staff ratios, whether in hospitals or external care facilities, medical personnel face the challenge of insufficient manpower Aurora Information has introduced the 'AU Care Recognition Service System', aiming to enhance medical completion rates through AI and 3D imaging technology Aurora Information introduces the 'AU Care Recognition Service System' using non-contact 3D imaging technology to instantly detect patient movements and physiological data from beds, significantly enhancing medical completion rates The 'AU Care Recognition Service System' incorporates Time-of-Flight ToF and mmWave Radar sensing technologies, coupled with point cloud and mmWave deep learning analysis AI algorithms, significantly strengthening computer vision and image processing capabilities Preserving Personal Privacy While Precisely Monitoring Patient Admission Status Aurora Information recently showcased the 'AU Care Recognition Service System' at the AI HUB conference With the use of Time-of-Flight ToF, mmWave Radar, and other sensing technologies, along with point cloud and mmWave deep learning analysis of AI algorithms, it greatly enhances computer vision and image processing capabilities Additionally, unlike traditional cameras that raise personal privacy concerns, this system uses non-contact 3D imaging technology for tracking bed exits and physiological data, providing instant notifications to users The system offers long-range, multi-target, high accuracy, continuity, and persistence advantages, all while maintaining patient privacy and precisely monitoring various admission statuses Through 3D imaging technology, when detecting potential danger actions like falls or tremors on the screen, it can immediately alert designated personnel or medical stations within 3 seconds, lowering the chances of sudden patient deaths The 'AU Care Recognition Service System' also measures heart rate, respiration, and other physiological data, using AI algorithms to predict the likelihood of sudden death 3D Imaging Technology for Immediate Reflection of Patient Emergency Situations How does it work in practice Suppose a patient falls out of bed when unattended, traditionally, they could only passively wait for help However, with the 'AU Care Recognition Service System' using 3D imaging technology to detect actions such as falls or tremors, it can immediately notify designated personnel or medical stations within 3 seconds, reducing the chances of sudden death by 20 it also measures heart rate, breathing, and other physiological data, using AI algorithms to assess the risk of sudden death, potentially reducing the death rate by over 40 Moreover, the system can simultaneously assess multiple patients, infants, and the elderly, enhancing timely care efficiency by more than 20, thereby significantly lightening the burden on medical staff Through the smart monitoring interface, medical personnel can clearly see real-time conditions in hospital wards or stairways, with intuitive icons helping them grasp the situation immediately Amidst the global challenges of aging and chronic diseases leading to high medical care costs, the 'AU Care Recognition Service System' effectively addresses the pain point of insufficient medical care According to the medical data released by the Ministry of Health and Welfare last year, if the 'AU Care Recognition Service System' were to be standardized, the market size for Taiwan's medical care institutions alone would be as high as 15 billion NTD Currently, Aurora Information is actively cooperating with the government, universities, and nonprofit organizations, hoping to contribute to medical care「Translated content is generated by ChatGPT and is for reference only Translation date：2024-05-19」

AI information technology can be utilized in very distinct industrial fieldsUnmanned factories, automated production lines, and robotic arms, etc To accelerate production lines to achieve economies of scale and simultaneously reduce costs, major production lines look forward to continuously optimizing in various aspects to reduce errors and increase yield rateIn the production line, quality control testing is a critical step that significantly impacts product quality Taiwan's Juou Technology Corporation stands out in its research and development efforts, not only addressing the management of Taiwan's resources and urban-rural development issues but also in its performance through innovative cross-disciplinary integrationWithin Juou's GEO IOT plan, two parts of the cross-disciplinary technology project are implementedPart oneWithin the AOI technology area, there is active investment in the technical development of big data, artificial intelligence, and data science It also involves the development of a high-speed, high-precision optical imaging inspection system in the production line quality management This technology utilizes machine vision with high focus and sensing to capture the surface images of products and automatically assess whether the products meet the quality standardsThis non-contact inspection tool not only speeds up the inspection process and identifies product defects but also automatically discriminates flawed products It is also applicable for the inspection of semi-finished products in the production line In enterprises that are major producers in Taiwan or those with a focus on quality, this optical combined with image processing system is necessaryPart twoBlood tests, crowd recognition, access control in danger areas construction sites, IVS, AIR vehicle controlBiometric recognition technology utilizes the characteristics of organisms for identification, including iris, retinal, and body shape recognition It also enables the monitoring of people flow within a specific space to quickly capture more information about crowd movements Beyond handling crowd recognition, Juou Technology also employs recognition technologies for control and monitoring of hazardous areas such as construction sites and production lines For instance, in spatial detection technology—if a construction worker forgets to wear a safety helmet or safety suit, the detection equipment will notify them upon entry to the site, thus reducing risks caused by human oversightCommitted to the innovation of the brand and the integration between sectors, it is believed that through digital technology and multi-disciplinary integration, the sustainable and innovative development of Taiwan will be driven「Translated content is generated by ChatGPT and is for reference only Translation date：2024-05-19」

For parents with young children, discussing matters with school teachers via the instant messaging app LINE is a good method Although LINE groups are convenient, if there are too many people talking at once, it can make what should be simple communication inefficient Starting from the service of campus smart vending machines, solving the pain points in parent-teacher communication In order to facilitate better communication between parents and schools, Chester International Co, Ltd and Zhonghua Lin Co, Ltd have collaborated to launch the SchoolBot service In fact, last year when Chester deployed the iMvending smart vending machine service at Taipei's Xingan Elementary School, they observed many communication pain points between parents and teachers This influenced the development of SchoolBot Utilizing the popular instant messaging app LINE as its platform, SchoolBot is essentially a robotic assistant for schools, incorporating LINE Bot and AI auto-response technology, and offering a pioneering one-on-one communication method that eliminates the clutter of traditional LINE groups SchoolBot incorporates LINE Bot and AI automatic response techniques, along with a pioneering one-on-one communication method in Taiwan, ensuring avoidance of the chaotic situation common in traditional LINE groups Automatic categorization makes it user-friendly, and privacy is secured Thanks to the adoption of AI technology, SchoolBot offers more applications than expected, providing a user-friendly and intuitive operation experience, as if chatting with a real person Whether helping children take leaves, sending messages to specific teachers, the system emphasizes categorization, allowing schools to identify and send messages to the appropriate recipients For parents, aside from basic inquiries, SchoolBot also supports categorization by child's student ID Upon the first use, when parents input their child's student ID, the system automatically categorizes it into the corresponding class, eliminating the need for manual selection and ensuring privacy security Whether helping children take leaves or sending messages to specific teachers, SchoolBot provides a smooth and intuitive operation experience, similar to chatting with a real person Additionally, SchoolBot caters to the common discussion scenarios faced by teachers and parents, integrating a practical online survey service This not only saves time and streamlines work processes for teachers but also supports the modern trend of going paperless Notably, the SchoolBot system automatically 'levels up' at the start of a new academic year, automatically advancing students to the next grade level, addressing the manual update problem that similar services face Chester has already collaborated with Xingan Elementary School, and hopes to expand the service into various community-based scenarios, creating a more convenient communication platform Besides schools, SchoolBot is also suitable for various scenarios like daycare centers, tutoring centers, and parent associations Chester has already partnered with Taipei's Xingan Elementary School and plans to continue expanding SchoolBot to other places, aiming to build a more convenient communication bridge for community-based scenarios「Translated content is generated by ChatGPT and is for reference only Translation date：2024-05-19」