Reinforced Labels: Multi-Agent Deep Reinforcement Learning for Point-Feature Label Placement

Bobák P., L. Čmolík and M. Čadík, Reinforced Labels: Multi-Agent Deep Reinforcement Learning for Point-Feature Label Placement, IEEE Transactions on Visualization and Computer Graphics, 30(9), pages 5908-5922, 2024.
DOI: 10.1109/TVCG.2023.3313729. 

Abstract

Over the recent years, Reinforcement Learning combined with Deep Learning techniques has successfully proven to solve complex problems in various domains, including robotics, self-driving cars, and finance. In this paper, we are introducing Reinforcement Learning (RL) to label placement, a complex task in data visualization that seeks optimal positioning for labels to avoid overlap and ensure legibility. Our novel point-feature label placement method utilizes Multi-Agent Deep Reinforcement Learning to learn the label placement strategy, the first machine-learning-driven labeling method, in contrast to the existing hand-crafted algorithms designed by human experts. To facilitate RL learning, we developed an environment where an agent acts as a proxy for a label, a short textual annotation that augments visualization. Our results show that the strategy trained by our method significantly outperforms the random strategy of an untrained agent and the compared methods designed by human experts in terms of completeness (i.e., the number of placed labels). The trade-off is increased computation time, making the proposed method slower than the compared methods. Nevertheless, our method is ideal for scenarios where the labeling can be computed in advance, and completeness is essential, such as cartographic maps, technical drawings, and medical atlases. Additionally, we conducted a user study to assess the perceived performance. The outcomes revealed that the participants considered the proposed method to be significantly better than the other examined methods. This indicates that the improved completeness is not just reflected in the quantitative metrics but also in the subjective evaluation by the participants. 

Preprint

Donload the preprint of our paper.

Supplementary material

For detailed description of the conducted experiments, please look at the supplementary material. 

Dataset

The dataset used in the experiments is available for download. Whenever you use the dataset, please acknowledge it by citing our paper. 

Acknowledgements

This work was supported by project LTAIZ19004 Deep-Learning Approach to Topographical Image Analysis; by the Ministry of Education, Youth and Sports of the Czech Republic within the activity INTER-EXCELENCE (LT), subactivity INTER-ACTION (LTA), ID: SMSM2019LTAIZ and by Grant Agency of CTU in Prague grant No. SGS22/173/OHK3/3T/13 - Research of Modern Computer Graphics Methods 2022-2024. Computational resources were mainly supplied by the project ”e-Infrastruktura CZ” (eINFRA CZ ID:90140) supported by the Ministry of Education, Youth and Sports of the Czech Republic.