Social connectedness is a significant factor in preventing cognitive decline and improving psychological well-being among older adults. However, older adults may suffer from social disengagement due to physical impairments and social loss. In an increasingly digital world, older adults also face challenges in building and maintaining social relationships due to low technological acceptance and a lack of appropriate tools. Personal narratives, telling stories about one’s past, contribute to cognitive and mental health among older adults. They boost self-confidence, enhance meaning of life, and have therapeutic value for those with mild cognitive impairment and dementia. Further, sharing biographical stories with others allows one to quickly build and deepen social connections, reducing stigma and fostering mutual understanding. This project aims to develop an age-friendly digital storytelling social platform supported by AI-enhanced speech and image recognition technologies. This innovative system enables older adults to capture and record their personal stories and experiences, transforming them into digital narratives that can be shared through an online platform across generations. A multilingual AI speech recognition system is embedded to ensure accurate and reliable transcription of the spoken words. The system also installs an AI image recognition system to automatically identify the location and other key information from the user-uploaded photo to accompany and enrich the storytelling.

Professor Helen Meng’s project on AI-powered solution for real-time, high-performance speech recognition, advanced machine translation and streamlined content summarization was awarded a gold medal. This advanced technology has empowered Radio Television Hong Kong (RTHK) to enhance its live broadcast of the Chief Executive’s 2024 Policy Address and the Financial Secretary’s 2024 Budget Speech. By leveraging cutting-edge speech recognition, the system transcribes spoken audio in real-time, while generative AI synthesizes the transcripts into concise, bullet-point highlights updated every minute. This innovation replaces the traditionally time-intensive processes of manual note-taking and summarization, delivering a streamlined and efficient solution. During live broadcasts on RTHK Channel 32, the AI-generated summaries are displayed as rolling bullet points, providing viewers with instant access to key insights and fostering a deeper understanding of these critical addresses.

The project by Professors Helen Meng and Xianming Gong, in collaboration with Professor Patrick Wong of the Brain and Mind Institute, designed a naturalistic, language-based fMRI task with AI technologies for detection and prediction of neurocognitive disorders and was awarded a silver medal.

Dongchao Yang, PhD advisee of Prof. Helen Meng and Prof. Xixin Wu, received one of only three Best Student Paper Awards for the entire conference.

Ms. Jincenzi Wu, PhD advisee of Professor Helen Meng, received the Area Chair Paper Award at ACL 2024. This paper was ranked within the top 1 % of the the entire conference.

Prof. Helen Meng was elected as a Distinguished Fellow of the Hong Kong Computer Society in recognition of her outstanding contributions.

Prof. Helen Meng and Prof. Xixin Wu were awarded a new project from the HKSAR Government's Quality Education Fund in May 2023. The project is titled "LingoTask: An AI-Powered English Teaching and Learning System" and aims to use AI technologies to enhance the teaching and learning of English as a second language.

Prof. Qi Dou, Prof. Helen Meng, et al. were awarded a new R&D grant in February 2023 from the Innovation and Technology Support Programme, in collaboration with the Dr. Stanley Ho Medical Development Foundation. The project is titled "AI-ENDO: AI-enabled Video Data Analytics to Improve Endoscopic Surgical Safety".

Prof. Xianmin Gong was awarded a grant from The Reconnect Travel Grant of the British Council to support his collaborative research with the University of Exeter and the University of Strathclyde, titled "How to motivate preparation for ageing? A cross-cultural evaluation on virtual reality age progression interventions in Hong Kong and Australia". Only four research teams in Hong Kong received this grant.

Prof. Helen Meng’s team was honored with the prestigious ISCA Best Student Paper Award at INTERSPEECH 2023 in Dublin. Their paper, titled “Diverse and Expressive Speech Prosody Prediction with Denoising Diffusion Probabilistic Model,” was recognized for its innovative methodology in leveraging speech analytics to detect early signs of cognitive decline. The award is presented by the International Speech Communication Association (ISCA), spotlighting the student-led research excellence accomplished under Prof. Meng’s leadership.

Dr. Thomas Lam Yuen Tung was awarded the Poster of Distinction at Digestive Disease Week (DDW) 2022. His poster, entitled “Artificial Intelligence-assisted diagnosis has more evidence to improve GI outcome than other specialties: an analysis of RCT from 2009-2021,” analyzed the impact of AI-assisted diagnosis in gastrointestinal medicine using data from randomized controlled trials spanning 2009 to 2021. This recognition by the American Gastroenterology Association highlights Dr. Lam’s contribution to advancing the role of AI in improving clinical outcomes in gastroenterology.

The research team led by Professor Helen Meng and Professor Xixin Wu, together with team members Kun Li, Tianhua Zhang, Liping Tang, Junan Li, and Hongyuan Li, was awarded First Place in the Doc2Dial@ACL 2022 Challenge. The team developed a multi-document, multi-domain automatic question-answering system that outperformed international competitors, achieving the top score on both official leaderboards. This achievement demonstrates CUHK’s leadership in the field of natural language processing and conversational AI.

The research team led by Prof. Helen Meng achieved first place in both the “TWO” task and the “CULTURE” task of the Affective Vocal Bursts Competition at the 10th International Conference on Affective Computing and Intelligent Interaction (ACII 2022). Their hierarchical regression chain framework outperformed global competitors in modeling non-linguistic vocalizations for emotion recognition. This result highlights BDDA’s expertise in developing advanced artificial intelligence for emotion understanding.

The research team led by Prof. Xixin Wu and Prof. Helen Meng, in collaboration with Tencent, achieved Second Place in the IEEE ICASSP 2022 Multi-Channel, Multi-Party Meeting Transcription (M2Met) Challenge. Their system, designed for speaker diarization, can automatically determine “who spoke when” in complex, multi-speaker audio environments. This strong result demonstrates the team’s advanced capabilities in speech technology and robust real-world performance in multi-party meeting transcription.

One of our Research Assistant Professor, Prof.  Xixin Wu’s paper titled “Exemplar‑based Emotive Speech Synthesis” was honored with the Best Paper Award at the Second Shenzhen Excellent Science & Technology Academic Paper Awards (深圳市第二届优秀科技学术论文) held in December 2022. The award recognized the significant advancements introduced by Wu’s work, which developed a novel approach for generating expressive speech by extracting emotion descriptors from exemplar audio and conditioning a neural TTS model, achieving substantial improvements in both emotional similarity and synthesized speech quality. The awards ceremony was co-organized by the Shenzhen Association for Artificial Intelligence and the Shenzhen Science & Technology Association, highlighting outstanding academic contributions within the region.

At the 2022 IEEE International Conference on Robotics and Biomimetics (IEEE ROBIO), Prof. Xixin Wu and Prof. Helen Meng, along with colleagues from The Chinese University of Hong Kong and The Hong Kong Polytechnic University, received the Best Paper Award in Robotics. Their paper, titled "Speech‑Vision Based Multi‑Modal AI Control of a Magnetic Anchored and Actuated Endoscope," presents a novel AI-driven system enabling surgeons to control an endoscopic camera via integrated speech commands and visual tracking. This system improves usability and autonomy in minimally invasive surgeries by intuitively aligning the endoscope’s view with surgical instruments.

Prof. Helen Meng’s team was awarded the Open Group Championship at the SciTech Challenge 2021 in Hong Kong, a prominent innovation and entrepreneurship competition organized by HKSTP. This accolade recognizes their leading-edge solution in the “Technology Enabled by Voice” category, showcasing her research group's strength in voice-based AI. The award highlights BDDA's continued excellence in applying AI to real-world technology innovation