![[2602.22703] Enhancing Geometric Perception in VLMs via Translator-Guided Reinforcement Learning](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.22703] Enhancing Geometric Perception in VLMs via Translator-Guided Reinforcement Learning
Summary
The paper presents GeoPerceive, a benchmark for evaluating geometric perception in vision-language models (VLMs), and introduces GeoDPO, a translator-guided reinforcement learning framework that significantly enhances VLMs' geometric reasoning capabilities.
Why It Matters
This research addresses the limitations of current VLMs in geometric reasoning, a critical aspect for applications in fields like robotics and computer vision. By introducing a novel benchmark and method, it paves the way for improved model performance and generalization in understanding complex visual data.
Key Takeaways
- GeoPerceive benchmark allows isolated evaluation of geometric perception in VLMs.
- GeoDPO framework utilizes a translator for enhanced reinforcement learning performance.
- Significant performance improvements observed: +26.5% in-domain and +39.0% on downstream reasoning tasks.
- Supervised fine-tuning may impair performance in out-of-domain scenarios.
- All codes are made publicly available for reproducibility.
Computer Science > Machine Learning arXiv:2602.22703 (cs) [Submitted on 26 Feb 2026] Title:Enhancing Geometric Perception in VLMs via Translator-Guided Reinforcement Learning Authors:Hao Yu, Shuning Jia, Guanghao Li, Wenhao Jiang, Chun Yuan View a PDF of the paper titled Enhancing Geometric Perception in VLMs via Translator-Guided Reinforcement Learning, by Hao Yu and 4 other authors View PDF Abstract:Vision-language models (VLMs) often struggle with geometric reasoning due to their limited perception of fundamental diagram elements. To tackle this challenge, we introduce GeoPerceive, a benchmark comprising diagram instances paired with domain-specific language (DSL) representations, along with an efficient automatic data generation pipeline. This design enables the isolated evaluation of geometric perception independently from reasoning. To exploit the data provided by GeoPerceive for enhancing the geometric perception capabilities of VLMs, we propose GeoDPO, a translator-guided reinforcement learning (RL) framework. GeoDPO employs an NL-to-DSL translator, which is trained on synthetic pairs generated by the data engine of GeoPerceive, to bridge natural language and DSL. This translator facilitates the computation of fine-grained, DSL-level scores, which serve as reward signals in reinforcement learning. We assess GeoDPO on both in-domain and out-of-domain datasets, spanning tasks in geometric perception as well as downstream reasoning. Experimental results demonstrate th...