AI_eng

Brief Description of Research


Recently, artificial intelligence (AI) has become a major research topic in various fields. There are more jobs that

require AI's related skills now. In view of this, a group of faculty members in our institute have been put together

to conduct research in big data analytics and machine learning (including deep learning) with applications in

multimedia signal processing (video summarization), social network analysis, recommender systems, and

precision marketing.

 

Smart Album System with Auto Captioner


 

The album can be organized “smartly’’ so that a picture can be searched by a keyword or a sentence!

DEEP LEARNING was used to automatically pick up the most important pictures and an improved LSTM was used to generate the semantic descriptions to boost user experiences.



 

Video Summarization and Captioning

A summarized video not only preserves the essential frames from the original video, but also captures their semantic information. With the proposed S2VT, SA, and data augmentation strategies, our approach further improves the performance of the semantic information. With the video summarization and captioning techniques, a user can easily seek any clips in a video by querying the semantic information!!



Face Hallucination for Very Low-Resolution Face Image

 

An effective and efficient identity-recognizable face hallucination technique for extreme low-resolution face images!

Left to right: low-res., interpolation, ours, and ground truths

 

Faculty members in the AI group

*Prof. Cheng-Shang ChangProf. Bor-Sen ChenProf. Chong-Yung ChiProf. Duan-Shin Lee

  Prof. Hwa-Chun LinProf. Ching-Te ChiuProf. Jen-Ming WuProf. Chia-Wen LinProf. Hsi-Pin Ma

  Prof. Yao-Win (Peter) HongProf. Che LinProf. Chih-Hao Huang

*Coordinator)

 

Projects

1. Project NameKnowledge Discovery and Its Applications of Large-scale Mobile Videos and Information for

                              5G Networks

  Project MembersProf. Cheng-Shang ChangProf. Chia-Wen LinProf. Chih-Hao Huang

                                   Prof. Shan-Hung WuProf. Duan-Shin LeeProf. Hwa-Chun Lin

                                   Prof. Hwann-Tzong ChenProf. Min SunProf. Hung-Min Sun

                                   Prof. Jung-Chun KaoProf. Che-Rung Lee

  Project PeriodNovember 2014 ~ July 2017

2. Project NameAnalysis of dynamic data linked by social networks

  Project MembersProf. Duan-Shin LeeProf. Cheng-Shang ChangProf. Wing-Kai Hon

  Project PeriodNovember 2014 ~ July 2017

3. Project NameLarge-Scale Cyber-Physical Data Analysis for Human-Centric Learning, Recommendation, and

                              Behavior Shaping

  Project MembersProf. Yao-Win (Peter) HongProf. I-Hsiang Wang (NTU)Prof. Chi-Chun (Jeremy) Lee

                                  Prof. Shih-Hau Fang (YZU)

  Project PeriodJune 2017 ~ May 2020

4. Project NameIntegrating deep learning, big data analytics, ChatBot, and customer relation management

                              systems for customer-centric precision marketing

  Project MembersProf. Che LinProf. Yun-Nung Chen (NTU)Prof. Galit Shmueli

                                   Prof. Li-Chen Cheng (SCU)

  Project PeriodJune 2017 ~ May 2020

 

Publications

1. Cheng-Shang Chag, Lee, Duan-Shin Lee, Li-Heng Liou, Sheng-Min Lu and Mu-Huan Wu, A probabilistic

    framework for structural analysis and community detection in directed networks, to appear in IEEE/ACM

    Transactions on Networking.

2. Chih-Chung Hsu and Chia-Wen Lin*, “CNN-based joint clustering and representation learning with feature

    drift compensation for large-scale image data,” IEEE Trans. Multimedia (Special Issue on Large-Scale

    Multimedia Data Retrieval, Classification, and Understanding), July, 2017.

3. Lorenzo Ferrari, Anna Scaglione, Reinhard Gentz, and Y.-W. Peter Hong, “Convergence results on pulse

    coupled oscillator protocols in locally connected networks,” to appear in IEEE/ACM Transactions on

    Networking.

4. C.-H. Lin, Chong-Yung Chi, Y.-H. Wang, and T.-H. Chan, “A fast hyperplane-based minimum-volume

    enclosing simplex algorithm for blind hyperspectral unmixing,” IEEE Trans. Signal Processing, vol. 64, no. 8,

    pp. 1946-1961, Apr. 2016.