My research interests broadly include pattern recognition (and computer vision), information retrieval (and data mining), and other related topics.
I am specifically interested in building applicable and novel technologies and systems for real classification, learning and retrieval applications, and have been developing several state-of-the-art and applicable technologies, e.g., Scene Text Detection and Recognition, Text Detection and Recognition from Web Pictures and Videos, Financial Document Image Analysis and Recognition, and Social Book (Product) Search and Recommendation.
I am also interested in the inter-discipline research between pattern recognition and information retrieval, e.g., biomedical figure search, and web video retrieval.
Currently, my main research projects include:
(1) iTERM: Intelligent TExt Reading in ubiquitous Multimedia (Robust Text Detection, Recognition, Retrieval and Mining in Natural Scenes, Web Images, Ubiquitous Documents, and Videos) (My team won the 1st place of "Text Localization in Real Scenes", "Text Localization in Born-Digital Images (Web and Email)" and "Text Segmentation in Born-Digital Images (Web and Email)" in ICDAR 2013 Robust Reading Competition, and further won the 1st place of ICDAR 2015 Robust Reading Competition "End-To-End Focused Scene Text Recognition (Generic)", "End-To-End Born-Digital Image Text Recognition (Generic)", "End-To-End Born-Digital Image Text Recognition (Weak)", and "Video Text Detection");
(2) iREBOOK: Intelligent social REtrieval for BOOKs ( My team won INEX 2014 “Social Book Search Track” (Suggestion Task)), information retrieval and recommendation techniques in social networks and biomedical literature (My team won several batches of 2015 BioASQ Challenge Task 3b: Biomedical Semantic QA.);
(3) iREVIDEO: Intelligent REtrieval for massive multimedia of web VIDEOs;
(4) iLEARN: Intelligent LEARNing in real application systems.