2D or 3D: Who Governs Salience in VLA Models? -- Tri-Stage Token Pruning Framework with Modality Salience Awareness
This work addresses acceleration needs for embodied intelligence systems using MVLA models, offering an incremental improvement over existing token pruning methods by incorporating modality-specific considerations.
The paper tackles the increased computational demand in multi-visual-modal Vision-Language-Action (MVLA) models by proposing a tri-stage token pruning framework that accounts for 2D/3D modality salience differences, achieving up to a 2.55x inference speedup with minimal accuracy loss and only 5.8% overhead.
Vision-Language-Action (VLA) models have emerged as the mainstream of embodied intelligence. Recent VLA models have expanded their input modalities from 2D-only to 2D+3D paradigms, forming multi-visual-modal VLA (MVLA) models. Despite achieving improved spatial perception, MVLA faces a greater acceleration demand due to the increased number of input tokens caused by modal expansion. Token pruning is an effective optimization methods tailored to MVLA models. However, existing token pruning schemes are designed for 2D-only VLA models, ignoring 2D/3D modality salience differences. In this paper, we follow the application process of multi-modal data in MVLA models and develop a tri-stage analysis to capture the discrepancy and dynamics of 2D/3D modality salience. Based on these, we propose a corresponding tri-stage token pruning framework for MVLA models to achieve optimal 2D/3D token selection and efficient pruning. Experiments show that our framework achieves up to a 2.55x inference speedup with minimal accuracy loss, while only costing 5.8% overhead. Our Code is coming soon.