Junyu Zhou is a graduate of the Department of Computer and Information Science (CIS) of the Faculty of Science and Technology (FST), whose real-time facial detection and expression recognition research earned him high praise in his final year project. His project utilized the FACS facial recognition coding system and demonstrated the feasibility of combining YOLO facial recognition and FACS, which can be applied in facial recognition fields such as automotive safety driving and criminal psychology. Zhou is grateful for the guidance and support provided by his advisor, Liming Zhang, and other professors in the department, and he feels fortunate to have found a highly applicable and innovative research direction. After graduation from UM, he plans to pursue further studies overseas and continue his research journey in the same field.
Choosing a Promising Discipline
Zhou, who hails from Xinjiang Uyghur Autonomous Region, has had a strong interest in numbers and computing since high school. In the current age of information technology, he believes that software engineering and related disciplines have a bright future with a wide range of applications, including Artificial Intelligence, big data, data analysis, software design and development. Universities in Hong Kong and Macau have a more international, open and independent academic environment, which has always been Zhou’s favorite place to study, because there are more academic exchanges and cooperation with overseas universities, which can provide students with a wider range of international learning and exchange opportunities . The Department of Computer and Information Science of the University of Macau ranks high in the world subject rankings and has a good reputation. Therefore, Zhou gave priority to the University of Macau when he selected university.
Enhancing Overall Capabilities
Zhou understands that university is not only for learning knowledge, but also a stage for accumulating diverse experiences and expanding networks. Looking back on his four years of meaningful university life, he believes that UM’s residential college system has allowed him to meet many like-minded friends, and participating in college activities has enriched his university experience. In terms of learning, Zhou took advantage of opportunities to actively study multiple non-professional subjects such as marketing and accounting, enhancing his overall qualities, cultivating cross-disciplinary thinking abilities, and laying a solid foundation for future development, hoping to better utilize his abilities in the workplace.
Overcoming Challenges
When discussing the most profound challenges he faced in university, Zhou cites his final year project as an example. The project used model-based recognition of facial expressions, which can be applied in identifying fatigued drivers in the automotive industry, assisting doctors in more accurately understanding a patient’s condition and psychological state in the medical field, or in the business industry for market research, advertising, and customer service. Initially, the project’s facial expression recognition accuracy was low due to insufficient data. Zhou and his teammate worked hard to search for more data to enhance the accuracy and recognition rate. With a stroke of insight, they decided to use facial expressions from virtual game characters as raw data to train the model’s recognition abilities, ultimately successfully obtaining a large amount of data and achieving ideal accuracy levels. This success greatly increased Zhou’s confidence and strengthened his determination to continue developing in this field. He is about to pursue a master’s degree at an overseas university. Although full of uncertainty and challenges, he believes that improving oneself, working hard, and persevering through difficulty will allow him to go further on his learning journey.
周俊宇是電腦及資訊科學系本科畢業生,憑著實時人臉檢測和表情識別的研究,在畢業設計項目中獲得優秀的評價。他的項目運用FACS人臉面部識別編碼系統,同時證明了yolo人臉識別和FACS同時結合的可行性,可應用在汽車安全駕駛和犯罪心理學等人臉識別。四年大學時光轉瞬即逝,除了感謝其指導教授張立明和學系其他教授的悉心指導和支持,也慶幸自己找到了應用性強且具創新性的研究方向。畢業後他將前往海外升學,精進此領域,繼續其科研探索之旅。
選擇具前景學科
周俊宇來自新疆維吾爾自治區,自中學就對數字及計算有濃厚興趣。他認為在當今資訊化時代,軟件工程的相關學科前景良好,有非常廣泛的應用,也可發展至多方領域如人工智慧、大數據、數據分析、軟件設計與開發等。港澳地區的大學擁有更國際化及開放自主的學術環境一直是周俊宇心之響往的升學地方,只因與海外大學有更多學術交流合作,可為學生提供更廣泛的國際化學習和交流機會。而澳門大學電腦及資訊科學系在世界學科排名較高,擁有較佳的聲譽,因此周俊宇大學升學時優先選擇澳門大學。
增強自身綜合能力
周俊宇深知大學不僅是學習知識,也是一個廣闊的舞台,通過多元化的經驗積累和人脈拓展,回顧四年充實有意義的大學時光,周俊宇認為大學住宿式書院制度令他認識不少志同道合的好友,一同參加書院活動更豐富了他的大學經歷。在學習上周俊宇抓緊機會,積極修讀多門非自身專業的學科如市場學及會計學,增強自身綜合素質,培養跨學科思維能力,也期望為未來發展打下堅實基礎,希望在職場上能更好地發揮自己的能力。
成功克服挑戰
談及大學中最深刻的挑戰,他以準備畢業設計項目為例,項目利用模型實識別時人臉和表情,可應用在識別疲勞轟駕駛、醫療領域以幫助醫生更精準地瞭解患者的病情和心理狀況或商業領域、包括市場調研、廣告投放和客戶服務等。項目最初因為數據不足而令模型的表情識別準確率較低,周俊宇與另一位組員歐陽堅程費煞思量, 苦尋更多數據以提升準確率及識別率。靈機一動下,他們決定使用虛擬遊戲中人物的表情作為原始數據以訓練模型的識別能力,最終成功把獲取大量數據並將準確率提升至理想水準。是次成功令周俊宇信心大增也更加堅定了在此領域發展的決心。未來他即將到海外高校修讀碩士課程,雖然充滿了不確定性和挑戰,但他相信,完善自我、努力不懈、不畏困難定能在學習路上越走越遠。