Search papers, labs, and topics across Lattice.
GUICrafter is a weakly-supervised GUI agent that utilizes massive unannotated screenshots to minimize reliance on costly human annotations, addressing the challenges of data scarcity in GUI agent development. The approach employs a two-stage curriculum learning framework, where the model first learns visual grounding from unannotated data and then refines its performance using a small set of high-quality annotations through reinforcement learning. Experimental results indicate that GUICrafter not only matches but often exceeds the performance of existing advanced systems like UI-TARS, while using merely 0.1% of the data required by those systems.
GUICrafter achieves superior GUI agent performance with just a fraction of the data, revolutionizing the way we think about training in data-scarce environments.
Data, as the fundamental substrate of modern intelligence, has greatly driven the development of current foundation models. Naturally, researchers aim to extend this paradigm to the domain of GUI agents, hoping to build strong GUI agents through a similar paradigm. However, GUI agent data cannot be directly harvested from the internet, making it costly and difficult to collect at scale. As a result, current GUI agents suffer from poor cross-device generalization and limited visual grounding ability for fine-grained GUI elements. As an attempt to address data challenge in GUI agents, we propose GUICrafter, a weakly-supervised GUI agent leveraging massive unannotated screenshots to substantially reduce the reliance on expensive human annotations. GUICrafter explores a curriculum learning framework for training GUI agents through two progressive stages. First, the model learns visual grounding from large-scale unannotated screenshots and webpages, leveraging the rich contextual signals inherent in GUI interactions without human annotations. Then, in Stage 2, we leverage a small amount of high-quality data to calibrate the model via reinforcement learning. Experiments show that GUICrafter achieves competitive, or even superior, performance to advanced systems like UI-TARS while using only 0.1% of its data. Furthermore, under the same amount of annotated data, GUICrafter surpasses all previous methods such as GUI-R1. Code, data, and models are available at https://github.com/fansunqi/GUICrafter.
Tsinghua AI