L62tnshszup7xcm9ypxl

姜宏昀 HUNG-YUN CHIANG

Student from NTHU, mainly research in NLP and Speech Processing with Deep Learning.

Competition

Text Labeling and Identification of Agricultural Articles ‑ AI CUP 2021, 7th / 402 teams

2021/09 - 2022/12, AICUP 2021 農業文章文字標註及辨識

The competition hope to use artificial intelligence to mark the results and build a text recognition analysis module to mark and retrieve relevant information from the agricultural text data, and to use it as a search application for pest and disease control technology.

Chinese Handwriting Recognition - E.SUN AI Challenge 2021 , 6th / 468 teams

2021/05 - 2021/07, 玉山人工智慧挑戰賽2021夏季賽 - 你,識字嗎?-中文手寫影像辨識

Automatic recognition of handwritten Chinese text in pictures through deep learning .

TSMC x Microsoft Careerhack 2020 - Mixed Reality Group , 2nd

2020/10 - 2021/01

We started from the user's point of view, clarified the user's needs, and thought about how to use HoloLens and MR concepts to solve problems. Using Unity as the development platform, we used the MRTK released by Microsoft as the tool to develop MR, and finally made the finished product and deployed it to HoloLens for testing usability.

AML Person Finding - E.SUN AI Challenge 2020 , 5th / 409 teams

2020/06 - 2020/08 , 玉山人工智慧公開挑戰賽2020夏季賽 - NLP應用挑戰賽

This is a part of competition using news to find possible (or used to) money laundering person. Finding these people can help Anti-money laundering (AML) in future. We used NER and Text Classification with BERT based model to solve this problem.

Formosa Speech Recognition Challenge 2020 -Taiwanese ASR , 10th 

2020/12 - 2021/01 , 台語語音辨識挑戰賽

Using Conformer as encoder and Self-Mixed Attention Decoder to improve performance  from Transformer architecture with ESPnet toolkit. 

Experience

Project Developer • Institute of Information Science, Academia Sinica 中研院資科所

2021/02 - 2021/08

  • Participate Conversational Open Domain Document-based Natural Speech Q&A System (COSQA) project, mainly research on information retrieval module.
  • Hung-Yun Chiang and Kuan-Yu Chen, "A BERT-based Siamese-structured Retrieval Model" ROCLING 2021: Conference on Computational Linguistics and Speech

Machine Learning Intern • E.SUN FHC 玉山金控

2020/11 - 2021/01

  • Take part in Text to Speech project, especially for reading and implementing paper, such as Glow-TTS, Flowtron, Mellotron etc on our own dataset.

Programming Intern • GIGABYTE 技嘉科技

2020/03 - 2020/09
  • I participated in three projects including optimized product search system with some NLP skills, backend website development and web crawler for collecting information of products from Yahoo, Momo and Shopee.

Education

National Tsing Hua University (NTHU) - Master

2021/09 - Current, overall GPA 4.3 / 4.3

CSIE, NLP Lab with advisor Prof. Jason S. Chang and co‑advised with Prof.Hong‑Han Shuai (NYCU EE)

  • Research on summarization.
  • Courses include Deep Learning, Data Mining, Social Network, Numerical Optimization

National Taiwan University of Science and Technology (NTUST) - Bachelor

2017/09 - 2021/06, CS major GPA 4.05 / 4.3, overall GPA 3.82 / 4.3

CSIE, NLP Lab with advisor Menphis Chen (Kuan-Yu Chen)

Skills


Project Experience

  • NLP
  • Text To Speech (TTS)
  • Speech Recognition
  • Web (Laravel, Angular)
  • APP (Swift)


Programming Language

  • Python / Pytorch
  • C++ / C
  • Javascript / HTML / CSS
  • Java
  • Php

Language

  • English - TOEIC 775
  • Chinese

Powered By CakeResume