Human in the Loop
Human-in-the-loop (HITL) systems integrate human expertise with artificial intelligence to improve the performance, robustness, and ethical considerations of AI systems. Current research focuses on optimizing HITL interactions across diverse applications, including autonomous driving, medical image analysis, and robotics, often employing techniques like active learning, reinforcement learning, and large language models to guide human input and improve model efficiency. The significance of HITL lies in its ability to address limitations of fully automated systems, particularly in complex or ambiguous tasks, leading to more accurate, reliable, and ethically sound AI solutions across various scientific and practical domains.
Papers
Demonstrating HumanTHOR: A Simulation Platform and Benchmark for Human-Robot Collaboration in a Shared Workspace
Chenxu Wang, Boyuan Du, Jiaxin Xu, Peiyan Li, Di Guo, Huaping Liu
I-MPN: Inductive Message Passing Network for Efficient Human-in-the-Loop Annotation of Mobile Eye Tracking Data
Hoang H. Le, Duy M. H. Nguyen, Omair Shahzad Bhatti, Laszlo Kopacsi, Thinh P. Ngo, Binh T. Nguyen, Michael Barz, Daniel Sonntag
A Versatile Framework for Analyzing Galaxy Image Data by Implanting Human-in-the-loop on a Large Vision Model
Mingxiang Fu, Yu Song, Jiameng Lv, Liang Cao, Peng Jia, Nan Li, Xiangru Li, Jifeng Liu, A-Li Luo, Bo Qiu, Shiyin Shen, Liangping Tu, Lili Wang, Shoulin Wei, Haifeng Yang, Zhenping Yi, Zhiqiang Zou
Challenging the Human-in-the-loop in Algorithmic Decision-making
Sebastian Tschiatschek, Eugenia Stamboliev, Timoth ee Schmude, Mark Coeckelbergh, Laura Koesten