analiza aiqbench
This research advances mobile AI by introducing Teach VLM, a model that translates visual screen trajectories into structured operational knowledge using vision-language models. The authors address data scarcity through a systematic flywheel approach and demonstrate superior performance in predicting action semantics compared to existing baselines. Ultimately, this work provides an interpretable framework for guiding downstream automation agents on heterogeneous mobile interfaces without relying solely on static UI perception.
GUITrans2Act: Understanding User Operational Behaviors from Mobile GUI Interactions with Vision-Language Models
arXiv:2606.12817v2 Announce Type: replace Abstract: Understanding the digital world on mobile devices is shifting from static UI perception to dynamic action comprehension. This capability enables models to convert visual state transitions into operational knowledge, defined as short natural-language sentences that describe action types, target UI elements, textual arguments, and execution orders. However, due to the highly diverse and heterogeneous UI designs across applications, existing vision-language models (VLMs) struggle to accurately infer these underlying operations. To bridge this gap, we introduce Teach VLM, a core model designed to translate mobile screen trajectories into step-wise operational knowledge by extracting and analyzing operation-related keyframes from demonstration videos. To address the scarcity of aligned training data, we develop a systematic data flywheel for scalable data acquisition. We further introduce a novel Chinese Mobile Screen Teach Benchmark for fine-grained evaluation. Building upon Teach VLM, we propose the Teach-and-Repeat paradigm, where the generated operational knowledge serves as an interpretable procedural reference to guide downstream screen-based execution agents. Extensive evaluations demonstrate that Teach VLM significantly outperforms strong VLM baselines, achieving state-of-the-art performance in operation semantics prediction. Furthermore, experiments in Android World show that our paradigm yields consistent Task Success Rate improvements for downstream agents. Together, Teach VLM and the Teach-and-Repeat paradigm offer a practical pathway from raw demonstrations to reusable task automation.
Authors:Yudong Zhang (1), Lei Hu (1), Daoyang Liu (2), Jiawei Liu (1), Yangfan Luo (1), Zhilin Gao (1), Zuojian Wang (1) ((1) Honor Device Co., Ltd, (2) The Chinese University of Hong Kong, Hong Kong, China)
Abstract:Understanding the digital world on mobile devices is shifting from static UI perception to dynamic action comprehension. This capability enables models to convert visual state transitions into operational knowledge, defined as short natural-language sentences that describe action types, target UI elements, textual arguments, and execution orders. However, due to the highly diverse and heterogeneous UI designs across applications, existing vision-language models (VLMs) struggle to accurately infer these underlying operations. To bridge this gap, we introduce Teach VLM, a core model designed to translate mobile screen trajectories into step-wise operational knowledge by extracting and analyzing operation-related keyframes from demonstration videos. To address the scarcity of aligned training data, we develop a systematic data flywheel for scalable data acquisition. We further introduce a novel Chinese Mobile Screen Teach Benchmark for fine-grained evaluation. Building upon Teach VLM, we propose the Teach-and-Repeat paradigm, where the generated operational knowledge serves as an interpretable procedural reference to guide downstream screen-based execution agents. Extensive evaluations demonstrate that Teach VLM significantly outperforms strong VLM baselines, achieving state-of-the-art performance in operation semantics prediction. Furthermore, experiments in Android World show that our paradigm yields consistent Task Success Rate improvements for downstream agents. Together, Teach VLM and the Teach-and-Repeat paradigm offer a practical pathway from raw demonstrations to reusable task automation.
20 pages, 9 figures. Yudong Zhang and Lei Hu contributed equally to this work. Zuojian Wang, and Zhilin Gao are corresponding authors
Artificial Intelligence (cs.AI)
arXiv:2606.12817 [cs.AI]
(or arXiv:2606.12817v2 [cs.AI] for this version)
Text integral sub licența deschisă a publicației · 279 cuvinte