Junhao Jia, Hanwen Zheng, Yueyi Wu, Huangwei Chen, Haishuai Wang, Jiajun Bu, Lei Wu. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Chart comprehension presents significant challenges for machine learning models due to the diverse and intricate shapes of charts. Existing multimodal methods often over-look these visual features or fail to integrate them effectively for Chart Question Answering. To address this, we introduce CHARTFORMER, a unified framework that enhances chart component recognition by accurately identifying and classifying components such as bars, lines, pies, titles, legends, and axes. Additionally, we propose a novel Question-guided Deformable Co-Attention (QDCAt) mechanism, which fuses chart features encoded by Chart-former with the given question, leveraging the question's guidance to ground the correct answer. Extensive experiments demonstrate a 3.2% improvement in mAP over the baselines for chart component recognition. For ChartQA and OpenCQA tasks, our approach achieves improvements of 15.4% in accuracy and 0.8 in BLEU score, respectively, underscoring the robustness of our solution for detailed visual data interpretation across various applications.11The source code and dataset are publicly available at https://github.com/VT-NLP/chartQA

Virtual maps for navigation are commonly categorized as either allocentric perspective maps or egocentric perspective maps, depending on the frame of reference. Route retracing means finding one’s way back from the destination to the origin. A 2 (navigation mode: allocentric/egocentric mode) × 2 (retracing mode: allocentric/egocentric mode) × 2 (gender) mixed design was implemented in this study to investigate how virtual map navigation modes affect route retracing performance in navigation. Participants’ performances in route retracing, ranked from the best to the worst, were in allocentric-allocentric, egocentric-allocentric, egocentric-egocentric, and allocentric-egocentric modes. The study demonstrated that the results mentioned above can be explained by the effect of the allocentric retracing mode and the consistency effect of the navigation-retracing mode. Females were more affected by these two effects than males, and gender differences disappeared in the allocentric-allocentric mode.