Search papers, labs, and topics across Lattice.
Libra-VLA, a novel Vision-Language-Action architecture, addresses the limitations of monolithic VLA models by introducing a coarse-to-fine dual-system that decomposes robotic manipulation into discrete macro-directional reaching and continuous micro-pose alignment. This decoupling balances learning complexity between a Semantic Planner (coarse intent) and an Action Refiner (precise alignment), leading to improved performance. Experiments demonstrate an inverted-U performance curve relative to action decomposition granularity, with peak performance achieved when learning difficulty is balanced between the subsystems.
Decomposing robotic manipulation into coarse and fine-grained actions isn't just conceptually cleaner鈥攊t actually unlocks a sweet spot where learning difficulty is balanced, boosting performance.
Vision-Language-Action (VLA) models are a promising paradigm for generalist robotic manipulation by grounding high-level semantic instructions into executable physical actions. However, prevailing approaches typically adopt a monolithic generation paradigm, directly mapping visual-linguistic features to high-frequency motor commands in a flat, non-hierarchical fashion. This strategy overlooks the inherent hierarchy of robotic manipulation, where complex actions can be naturally modeled in a Hybrid Action Space, decomposing into discrete macro-directional reaching and continuous micro-pose alignment, severely widening the semantic-actuation gap and imposing a heavy representational burden on grounding high-level semantics to continuous actions. To address this, we introduce Libra-VLA, a novel Coarse-to-Fine Dual-System VLA architecture. We explicitly decouple the learning complexity into a coarse-to-fine hierarchy to strike a training equilibrium, while simultaneously leveraging this structural modularity to implement an asynchronous execution strategy. The Semantic Planner predicts discrete action tokens capturing macro-directional intent, while the Action Refiner conditions on coarse intent to generate high-frequency continuous actions for precise alignment. Crucially, our empirical analysis reveals that performance follows an inverted-U curve relative to action decomposition granularity, peaking exactly when the learning difficulty is balanced between the two sub-systems. With the asynchronous design, our approach offers a scalable, robust, and responsive solution for open-world manipulation.
