Masked Facial Recognition
Zhang Yuwei, an 2021/2022 outstanding graduate of the Department of Computer and Information Science of FST, won the “Best Final Year Project Award” for her research, on Masked Facial Recognition. As covid-19 widespread across the world, wearing mask becomes a necessity in people’s daily life. However, due to the lack of related data, wearing a mask results in ineffectiveness of most current facial recognition systems. To address this problem, we propose a crop-based recognition method for masked facial recognition, which extracts features from cropped upper face image with a modified ResNet-18 and from face landmark coordinates with a perceptron. Extracted features are combines together to determine the prediction result. Extensive experiments carried out on 4 masked face recognition datasets prove the effectiveness of the proposed method on identifying individuals wearing mask.
UM’s learning environment
Regarding winning the Best Final Year Project Award, Zhang Yuwei feels excited by winning the award, she said that the reason is because her topic was close to the current situation which responses to the actual need during the epidemic, and the article used a lot of experiments and datasets to verify. Looking back to the four years at UM, Zhang Yuwei shared that UM’s English teaching environment is able to meet the standard of foreign universities, and the academic environment is positive to her. The university also has sufficient learning resources to assist students for broader exploration, and the quiet and beautiful campus also make Zhang Yuwei feel very comfortable, so she is able to concentrate on her study without distractions.
Receiving an offer from Yale University
Zhang Yuwei said that the professors of the college are also very friendly and put a lot of effort to guide students’ research. Zhang Yuwei, who has shown a strong interest in computer science in high school, aims to become a software engineer. Now she received an offer from Yale University for a master degree. She encourages students who are interested in pursuing a master degree could find out their interests and plan for their future ahead so as to achieve more and better opportunities for themselves.
張羽惟是本學年電腦及資訊科學系優秀本科畢業生，憑藉人臉口罩的識別與檢測，奪得「最優秀畢業設計項目獎」。隨著新型冠狀病毒在世界廣泛傳播，戴口罩已成為人們的新常態。然而，由於缺乏人臉口罩相對應的數據，目前戴口罩仍會導致大多數人臉識別與檢測系統無效。為解決這個難題，張羽惟提出了一種基於剪裁蒙版面部的識別方法，此方法為用改進的 ResNet-18 從上半部的人臉上提取特徵，並使用感知器從面部標誌坐標中提取特徵。 將提取的特徵組合在一起以確定預測結果。在張羽惟的研究中，在4組戴口罩人臉識別數據集上進行大量實驗，結果證明了所提出的方法在識別人臉口罩上是有效的。