OTTER: A Vision-Language-Action Model with Text-Aware Visual Feature Extraction
This addresses the challenge of maintaining pre-trained vision-language alignments in robotic action prediction, enabling better generalization to novel objects and environments.
The paper tackles the problem of preserving pre-trained semantic alignments in Vision-Language-Action models by proposing OTTER, which uses text-aware visual feature extraction to selectively pass task-relevant features to the policy, resulting in significant outperformance over existing models in simulation and real-world experiments with strong zero-shot generalization.
Vision-Language-Action (VLA) models aim to predict robotic actions based on visual observations and language instructions. Existing approaches require fine-tuning pre-trained visionlanguage models (VLMs) as visual and language features are independently fed into downstream policies, degrading the pre-trained semantic alignments. We propose OTTER, a novel VLA architecture that leverages these existing alignments through explicit, text-aware visual feature extraction. Instead of processing all visual features, OTTER selectively extracts and passes only task-relevant visual features that are semantically aligned with the language instruction to the policy transformer. This allows OTTER to keep the pre-trained vision-language encoders frozen. Thereby, OTTER preserves and utilizes the rich semantic understanding learned from large-scale pre-training, enabling strong zero-shot generalization capabilities. In simulation and real-world experiments, OTTER significantly outperforms existing VLA models, demonstrating strong zeroshot generalization to novel objects and environments. Video, code, checkpoints, and dataset: https://ottervla.github.io/.