Taiwan Patent on AI Pathological Voice Detection System
AINTU project principal investigator Professor Shih-Hau Fang obtains a Taiwan patent on "AI Pathological Voice Detection System." The system has collected over 1000 patient voices. By analyzing them using big data analytics, the system can identify 8 symptoms including vocal cord cancer, polyp, cyst and nodule while the trial subject needs to keep vocalizing to a microphone for only 1 second.
Finalist (Top 3) in 2018 IEEE Video and Image Processing (VIP) Cup
AINTU project principal investigator Professor Winston Hsu led his team of students in the finalist (top 3) for "Lung Cancer Radiomics -- Tumor Region Segmentation" in 2018 IEEE Signal Processing Society Video and Image Processing (VIP) Cup.
More than 80 Published Papers
AINTU project teams have published more than 80 papers in AAAI, IJCAI, CVPR, etc.
Share 2 AI Tools to the World
AINTU Co-Pi Research Fellow Ku Lun-wei and director Prof. Hsin-Hsi Chen have shared one AI tool respectively for AI workers worldwide:
1. CSentiPackage：Chinese Setiment analysis Package
2. CISA： Chinese Information Structure Analysis for Scientific Writing
World's #3 in CVPR 2018 Video Recognition Challenge — Moments in Time
World's #2 in Robust Vision Challenge 2018 — Instance Segmentation
#3 in Semantic Evaluation 2018 — Capturing Discriminative Attributes Task
World's #2 in DeepGlobe Workshop (Land Cover Classification)
AINTU project principal investigator Professor Frank Wang has won the second prize in the “DeepGlobe Workshop (Land Cover Classification)” CVPR 2018 Competition held at the world's premier computer vision conference.
Taiwan's #1 in Standard Chartered Financial Technology Campus Creative Challenge
AINTU diretor Professor Chen led the Natural Language Processing lab team of Chen Chung-ji, Huang Bo-cheng, Lin Yu-rou, Yang Zong-han, and Tsai Chia-wen and won the first place in the Standard Chartered Financial Technology Campus Creative Challenge.
World's #1 in Disguised Faces in the Wild
AINTU project principal investigator Professor Winston Hsu has won the top prize in the “Disguised Faces in the Wild” CVPR 2018 Competition held at the world's premier computer vision conference.
SSR-Net for Age Estimation (IJCAI 2018)
AINTU project co-principal investigator Professor Chuang Yung-yu has developed a new face image age estimation model, with an error of 3.16 years, and a gap of 0.5 years from the best model. The parameter is less than 1/1000, and the entire model parameter is only 0.3MB, making it ideal for embedded systems. The model was published in the top AI conference IJCAI 2018.
World's First Completely Unsupervised Voice Recognition System
AINTU project co-principal investigator Professor Lee Hung-yi has unveiled the world’s first speech recognition system that is completely trained in an unsupervised manner, allowing the machine to learn language B from language A on its own!
Share 3 ML Datasets to the World
AINTU project PI Professor Winston Hsu and co-PI Professor Lee Hung-yi have shared data sets (two from Hsu, one from Lee) for AI workers worldwide.
1. DroneViewBI: Drone-View Building Identification
2. NetiLook: Netizen-Style Commenting on Fashion Photos4, Country Style Clothes Photo Chat Robot
3. A New Listening Comprehension Task: SpokenSQuAD Listening Comprehension