Document Understanding
Deep research on multimodal document understanding, generation, and intelligent analysis
H!NT Lab
Human-Intelligence iNTegration Lab
Document Understanding · Human-Computer Interaction · Collaborative Cognition
H!NT Lab (Human-Intelligence iNTegration Lab) focuses on exploring cutting-edge technologies for the deep integration of human intelligence and machine intelligence, dedicated to building more natural, efficient, and intelligent human-AI collaboration systems.
Research Areas
Our research centers on three core areas:
Human-Computer Interaction
Developing more natural and efficient interaction technologies and human-AI collaboration systems
Collaborative Cognition
Exploring synergistic mechanisms between human cognition and AI in complex tasks
These three directions mutually support and synergistically develop: document understanding provides the core information processing capability, advanced interaction technologies facilitate smoother communication, and collaborative cognition research fundamentally optimizes the overall intelligence of human-AI symbiotic systems.
Research Projects
Major research projects ongoing in our lab
DEEPLING
Deepling.tech
An AI-powered tech commentary self-media with a distinct personality, fully automated and known for its sharp, witty tech reviews.
DeepTrans
DeepTrans Studio
Professional deep translation studio. An AI-powered high-quality translation platform supporting intelligent translation and post-editing of multilingual documents.
DeepMed
DeepMed Search
Intelligent medical literature retrieval system. Utilizing deep learning and knowledge graph technology to provide precise literature search and recommendation for medical researchers.
CLT
ChineseLawTranslation
Chinese law bilingual translation project. Building a high-quality Chinese-English legal terminology database to promote international dissemination of Chinese law.
Publications
Selected research publications from our lab
2026
QAlign-RAG: Causal-Aware Pseudo-Question Generation and Debate-Driven Verification for Medical RAG
KSEM
2026
HeterMV: Multi-view reasoning over source-aware heterogeneous evidence graph for multi-source fact verification
IPM
2026
MoE-TextDiffuser: Fine-grained control of font and color in visual text rendering
NEUCO
2025
Mixed Information Bottleneck for Location Metonymy Resolution Using Pre-trained Language Models
TALLIP
Contact
Questions about the lab, research collaboration, or student opportunities.