即時手寫辨識於視訊會議平台

Real-time Handwritten Machine Learning Recognition Applied on a Video Call Platform

范宇茗(Fabriciano Carrasco Espejo)

指導教授 ： 林宗男

國立臺灣大學/電機資訊學院/電信工程學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202104371

摘要

none

關鍵字

網頁即時通訊 ； 手寫辨識 ； 實時流式傳輸 ； 視訊會議 ； AR/VR

並列摘要

Due to the pandemic, schools and universities have changed their teaching methodology and have adopted the online classroom as a teaching format. However, the quick move to online teaching produced an ecosystem of online classroom platforms with a lack of features to fulfill the needs for different teaching situations. A common problem that has arisen during online classes is the lack of bandwidth for the increasing amount of users connected simultaneously. This has created situations where the presenter wants to share his slides for the attendees using a screen share feature, but the bandwidth is not enough and the image is not clear and the characters are illegible. This situation is accentuated when the presenter wants to ”handwrite” on the slides, using hardware that is not specifically designed for that function such as a computer mouse. Therefore, in this work, I propose an online teaching platform that usesWebRTC to make the application accessible for everyone from a browser, a shared blackboard with a Machine Learning algorithm that runs locally in the browser to recognize the handwritten characters and share the slides, everything in a 3D VR environment to enhance the experience of the attendees.

並列關鍵字

WebRTC ； Handwriting Recognition ； Real-time Streaming ； Video Conference ； AR/VR

參考文獻

R. Ubell, “How online learning kept higher ed open during the coronavirus crisis.”https://spectrum.ieee.org/tech-talk/at-work/education/how-online-learningkept-higher-ed-open-during-the-coronavirus-crisis, May 2020. Accessed: 2021-10-05.

K. Blessing and X. Qian, “Evaluation of online pedagogy among higher education international students in china during the covid-19 outbreak,” Creative Education, vol. 11, pp. 2262–2279, Nov 2020.

S. Saha, “A comprehensive guide to convolutional neural networks.”https://towardsdatascience.com/a-comprehensive-guide-to-convolutional-neuralnetworks-the-eli5-way-3bd2b1164a53, Dec 2018. Accessed: 2021-10-05.

S. Team, “Convolutional neural networks (cnn): Step 3 - flattening.”https://www.superdatascience.com/blogs/convolutional-neural-networks-cnnstep-3-flattening, Aug 2018. Accessed: 2021-10-05.

E. Crew, “Vr tracking: how does it really work?: our blog.”https://4experience.co/vr-tracking-meet-degrees-of-freedom/, Nov 2019. Accessed:2021-10-05.