UIShift: Enhancing VLM-based GUI Agents through Self-supervised Reinforcement Learning
By: Longxi Gao, Li Zhang, Mengwei Xu
Potential Business Impact:
Teaches computers to understand app screens without human help.
Training effective Vision Language Models (VLMs) for GUI agents typically relies on supervised fine-tuning (SFT) over large-scale annotated datasets, where the collection process is labor-intensive and error-prone. In this work, we propose a self-supervised inverse dynamics task to enable VLMs to learn from GUI transition pairs by inferring the action that caused that transition. This training task offers two advantages: (1) It enables VLMs to ignore variations unrelated to user actions (e.g., background refreshes, ads) and to focus on true affordances such as buttons and input fields within complex GUIs. (2) The training data can be easily obtained from existing GUI trajectories without requiring human annotation, and it can be easily scaled through automatic offline exploration. Using this training task, we propose UI-shift, a framework for enhancing VLM-based GUI agents through self-supervised reinforcement learning (RL). With only 2K training samples sourced from existing datasets, two VLMs -- Qwen2.5-VL-3B and Qwen2.5-VL-7B -- trained with UI-Shift achieve competitive or superior performance on grounding tasks (ScreenSpot-series benchmarks) and GUI automation tasks (AndroidControl), compared to SFT baselines and GUI-specific models that explicitly elicit reasoning abilities during RL. Our findings suggest a potential direction for enhancing VLMs for GUI agents by leveraging more self-supervised training data in the future. Code, model, and data are available at: https://github.com/UbiquitousLearning/UIShift
Similar Papers
VLM Q-Learning: Aligning Vision-Language Models for Interactive Decision-Making
Machine Learning (CS)
Teaches computers to see and follow instructions.
Enhancing Visual Grounding for GUI Agents via Self-Evolutionary Reinforcement Learning
Artificial Intelligence
Helps computers understand and click on screen buttons.
ZeroGUI: Automating Online GUI Learning at Zero Human Cost
Artificial Intelligence
Teaches computers to use apps by themselves.