MAIN-VLA: Modeling Abstraction of Intention and eNvironment for Vision-Language-Action Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917242114932736 |
|---|---|
| author | Zhou, Zheyuan Du, Liang Sun, Zixun Zhou, Xiaoyu Ye, Ruimin Chen, Qihao Chen, Yinda Qiu, Lemiao |
| author_facet | Zhou, Zheyuan Du, Liang Sun, Zixun Zhou, Xiaoyu Ye, Ruimin Chen, Qihao Chen, Yinda Qiu, Lemiao |
| contents | Despite significant progress in Visual-Language-Action (VLA), in highly complex and dynamic environments that involve real-time unpredictable interactions (such as 3D open worlds and large-scale PvP games), existing approaches remain inefficient at extracting action-critical signals from redundant sensor streams. To tackle this, we introduce MAIN-VLA, a framework that explicitly Models the Abstraction of Intention and eNvironment to ground decision-making in deep semantic alignment rather than superficial pattern matching. Specifically, our Intention Abstraction (IA) extracts verbose linguistic instructions and their associated reasoning into compact, explicit semantic primitives, while the Environment Semantics Abstraction (ESA) projects overwhelming visual streams into a structured, topological affordance representation. Furthermore, aligning these two abstract modalities induces an emergent attention-concentration effect, enabling a parameter-free token-pruning strategy that filters out perceptual redundancy without degrading performance. Extensive experiments in open-world Minecraft and large-scale PvP environments (Game for Peace and Valorant) demonstrate that MAIN-VLA sets a new state-of-the-art, which achieves superior decision quality, stronger generalization, and cutting-edge inference efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_02212 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MAIN-VLA: Modeling Abstraction of Intention and eNvironment for Vision-Language-Action Models Zhou, Zheyuan Du, Liang Sun, Zixun Zhou, Xiaoyu Ye, Ruimin Chen, Qihao Chen, Yinda Qiu, Lemiao Computer Vision and Pattern Recognition Despite significant progress in Visual-Language-Action (VLA), in highly complex and dynamic environments that involve real-time unpredictable interactions (such as 3D open worlds and large-scale PvP games), existing approaches remain inefficient at extracting action-critical signals from redundant sensor streams. To tackle this, we introduce MAIN-VLA, a framework that explicitly Models the Abstraction of Intention and eNvironment to ground decision-making in deep semantic alignment rather than superficial pattern matching. Specifically, our Intention Abstraction (IA) extracts verbose linguistic instructions and their associated reasoning into compact, explicit semantic primitives, while the Environment Semantics Abstraction (ESA) projects overwhelming visual streams into a structured, topological affordance representation. Furthermore, aligning these two abstract modalities induces an emergent attention-concentration effect, enabling a parameter-free token-pruning strategy that filters out perceptual redundancy without degrading performance. Extensive experiments in open-world Minecraft and large-scale PvP environments (Game for Peace and Valorant) demonstrate that MAIN-VLA sets a new state-of-the-art, which achieves superior decision quality, stronger generalization, and cutting-edge inference efficiency. |
| title | MAIN-VLA: Modeling Abstraction of Intention and eNvironment for Vision-Language-Action Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.02212 |