Despite the countless tele-therapy apps, a huge segment of the population has been neglected: employees. According to a survey in 2020, 27% of the Hong Kong workforce is suffering from a mental health problem. We knew we had to solve this. Introducing mindAId, which enables employers to provide employees with necessary mental health resources such as sessions with licensed therapists with easier scheduling, in-office support groups, AI chatbot and data-driven metrics. Hence, providing the much-needed support to employees and boosting their productivity, while employers simultaneously meet their SDG goals
In the digital era, Classical Music plays a dominant role in our daily lives. It often concurs in commercials, games, movies, and more. In general, Classical Music falls between the Baroque, Classical, Romantic, and Modernist eras. Among all these different musical periods, musicians nowadays might be proficient only to a number of them. With the AI, composers, musicians, or even non-specialists without Classical Music backgrounds could quickly get on to their journey in Classical Music composition. Recently, AI music composition has become more popular and accessible to people. However, most AI music generator platforms only focus on creating modern Music. Moreover, they usually categorize all genres of Classical Music into a single category – Classical Music. Since most research in this field adopted LSTMs to generate Music, it is worthwhile to discover different types of machine learning methods such as CNN-GANs for this task. Perhaps the proposed models could set successful precedence in this research area. This project aims to generate Classical Music using generative models – Bi-LSTM and CNN-GAN to compose Classical Music for four Classical Music genres and evaluate their performance respectively and collectively. Also, to further explore and strengthen the area of AI in music composition.
ReHertz is an online conferencing tool for real-time music making and audio production collaboration. Our mission is to bring sounds together wherever and whenever with technology. The pandemic has made it difficult for people to make music together face-to-face. Many musical groups have resorted to practising online through video conferencing apps, but delays cause synchronisation problems. ReHertz solves this problem by synchronising audio streams and matching the sounds in phase. ReHertz synchronizes and processes multiple inputs in real time to help our users produce the best music by overcoming latency issues. Performers will first sing individually to the backing track, and the streams will be sent from their computers to the Conductor’s computer. ReHertz will then synchronize the streams from different performers and play it back to the conductor within 1-2s. The conductor could then set the synchronized stream as the backing track for the performers, allowing performers to sing along with the whole choir’s latest recording. This way, the performers can listen to other’s voices and sing in real-time even without high-speed internet connections, recreating the experience of group music making instead of only hearing their own voices, as with current software solutions, revolutionising online music collaboration.
With the increasing number of bus accidents in recent years (For example, the Tai Po Road accident in 2018, a bus crashing into a tree in Fanling Highway in late 2019), the public is getting more concerned about the safety of franchised buses. Since most of these accidents are caused by speeding, some relevant safety devices have to be introduced. Now it is a demonstration to regulate the speed of a bus at a bus model. A sensor was installed at the front wheel of the bus model. The sensor will measure the speed of revolution of the wheel and then the measured value will convert into a speed unit by simple programming. When the speed exceeds 50km/h (all these speeds are scaled with the model unit), the safety device will ring for a small while. When the speed exceeds 70km/h the safety device will ring continuously until the speed drops below 70km/h (In Hong Kong, the maximum speed limit of a franchised bus is 70km/h). It is believed that it may remind drivers not to drive too fast, hope this may reduce the risk of traffic accident.
In today’s world, designing smart cities require multifaceted solutions for complex problems that require a lot of systems to work together in an integrated way – and that can be very difficult to conceptualise, let alone solve with the help of technology. This where the Internet of Things (IoT) and integrative systems with smart sensors have a vital role to play. However, one big pain point of creating these IoT systems is the smooth interaction between traditional processes and the technology innovation while having the human factors present. Keeping these in mind, this project is proposing a scalable, cloud-based multi-platform autonomous ecosystem that can monitor, control, assess and authenticate several devices with multiple sensors from both consumer and commercial end points. This is a multi-platform solution with a mobile app, a web admin portal as well as a proprietary embedded system that can bring multiple complex sensor data under one umbrella and let multiple stakeholders interact with them through simple, customisable human-centered interface designs. As a use case, this project will focus on designing proposed cloud-based platform customised for a wastewater treatment system however, this setup can be scaled further on for any multifaceted IoT device.
In the field of stock investment, decision-making depends on sophisticated information processing and data analysis. However, limitations of expertise, time, and resources make retail investors suffer from information overload and information imbalance. Recent improvement in computing power and the availability of high-quality data empower us to leverage Deep Learning techniques to bridge the gap. This project presents an integrated system that comprehensively monitors the risks of individual stocks and the overall market. The stock risk is estimated based on quantitative data of related stocks, where the relationship between business is measured by constructing an Enterprise Knowledge Graph using public knowledge. On the other hand, the market risk is estimated based on daily news. For each risk, a Temporal Convolutional Network is trained to output a continuous risk level that reveals both direction and amplitude of incoming changes. Eventually, key information and the predicted risk levels are organized into a condensed dashboard to interact with retail investors. Experiments on focal stocks in U.S. market suggest 67% and 56% accuracy in stock risk and market risk modeling respectively. Besides, visualizations on testing data show that our model has the potential to inform reverse changes of a stock movement ten days in advance.
Since electricity cannot be stored in a large amount, the supply and demand have to be well balanced. As a result, short term load forecast (STLF) would be essential for the power industry to allocate generation resources, spinning reserve, performing system maintenance etc. The objective of the project is to examine the feasibility of machine learning in short term load forecasting. As the electrical loading data in Hong Kong is not opened to the public, the data from Singapore Energy Markey Authority is being used instead. Historical load profile, calendar data and weather data would be used for the forecast. As historical weather forecast is not available, it cannot account for the weather forecast error into the STLF results. Scikit-learn would be used as the tool for machine learning with python as the programming language. Application of such tool would be to provide a more accurate and easy way to perform short term load forecast. If accurate load forecast is available, the power company can better allocate its resources such as lowering the extra generation capacity required to cater forecast error and improve grid security when the forecasted peak load is too high by using the day-ahead demand response scheme.
Jerkin Psalm gives a brand-new meaning to shopping clothes online. You can put on up-to-the-minute fashion items using extended reality technology—- without actually putting on them. So convenient is it that you can save your precious time for an important date, in your perfect outfit from Jerkin Psalm. Wonder if that bold crop top fits you? Changing styles to JK girls or 80’s extravagant attire? Just try on without getting embarrassed in boutiques. In view of Covid 19, this system can also prevent direct contact with the clothes to ensure hygiene level. Don’t need to worry about the size because the system will find you the best-fit item by detecting your body size. So you can just choose whatever you want to wear in little time. Not only can you try on the costume virtually, but also make the order directly from the shops, ranging widely from Uniqlo, H&M to Alexander McQueen and Balenciaga. Online shopping is that simple and divine
In modern society, smart and green is of paramount importance in forming an ideal city, which the government had put tremendous effort into it. For the sake of greater resource allocation, we believed that new technology such as Ai and machine learning combining with sound processing could provide a solution for traffic route planning such as bus route and mtr route. Also, with a mature base of data and continuous AI algorithm developent, the AI system could be evolved to analyse the meaning of different sound level, with may aid tech-based marketing and thus obtain valuable data. The business model will be mainly selling hardware of sound collecting device in stage one and selling data to org or government in stage two.
Today, computer manufacturers are trying to enhance chip’s function to reduce the weight of computer. Actually, we can change its shape to make it more convenient and easier to carry. Intelligent scroll looks like the Chinese traditional scroll will redefine the personal computer of the future. It also has two shaft levers on two sides to fix the scroll screen.
In industrial design, we applied flexible OELD transparent scroll screen, Dolby surround sound and intelligent pencil (inside the shaft lever). Equipping with 30000mAh battery, the device endurance time can reach 35 hours.
In system interaction, the intelligent scroll will be equipped with Link OS system which can be compatible with Windows and Mac OS system. Specifically, in addition to traditional voice assistant, eye assistant can also be applied to control the scroll computer.
In smart future, we will apply the newest L1 processor chip which integrate GPU and CPU to minimize the space occupied by hardware. Intelligent scroll will adopt cloud computer model and use computing power in the cloud through 5G.
Intelligent scroll will mainly be used in the construction industry. Workers can use it to design, construct, check process and allocate resources.