# Look, Zoom, Understand: The Robotic Eyeball for Embodied Perception

arXiv: https://arxiv.org/abs/2511.15279 · HTML: https://mlsanity.com/paper/2511.15279/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 32 references that are in the map.
- First posted on arXiv 19 Nov 2025.

## Builds on (most cited first, 21 of 21)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [OpenVLA: An Open-Source Vision-Language-Action Model](https://mlsanity.com/paper/2406.09246/) (2024)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [CLIPort: What and Where Pathways for Robotic Manipulation](https://mlsanity.com/paper/2109.12098/) (2021)
- [An Embodied Generalist Agent in 3D World](https://mlsanity.com/paper/2311.12871/) (2023)
- [Vision-Language Foundation Models as Effective Robot Imitators](https://mlsanity.com/paper/2311.01378/) (2023)
- [TinyVLA: Towards Fast, Data-Efficient Vision-Language-Action Models for Robotic Manipulation](https://mlsanity.com/paper/2409.12514/) (2024)
- [Open-World Object Manipulation using Pre-trained Vision-Language Models](https://mlsanity.com/paper/2303.00905/) (2023)
- [Grounding DINO 1.5: Advance the "Edge" of Open-Set Object Detection](https://mlsanity.com/paper/2405.10300/) (2024)
- [RoboMamba: Efficient Vision-Language-Action Model for Robotic Reasoning and Manipulation](https://mlsanity.com/paper/2406.04339/) (2024)
- [EfficientVLA: Training-Free Acceleration and Compression for Vision-Language-Action Models](https://mlsanity.com/paper/2506.10100/) (2025)
- [EvaGaussians: Event Stream Assisted Gaussian Splatting from Blurry Images](https://mlsanity.com/paper/2405.20224/) (2024)
- [Move to Understand a 3D Scene: Bridging Visual Grounding and Exploration for Efficient and Versatile Embodied Navigation](https://mlsanity.com/paper/2507.04047/) (2025)
- [VLA-0: Building State-of-the-Art VLAs with Zero Modification](https://mlsanity.com/paper/2510.13054/) (2025)
- [Vision in Action: Learning Active Perception from Human Demonstrations](https://mlsanity.com/paper/2506.15666/) (2025)
- [Active Vision Might Be All You Need: Exploring Active Vision in Bimanual Robotic Manipulation](https://mlsanity.com/paper/2409.17435/) (2024)
- [Uncertainty-Informed Active Perception for Open Vocabulary Object Goal Navigation](https://mlsanity.com/paper/2506.13367/) (2025)
- [AVR: Active Vision-Driven Precise Robot Manipulation with Viewpoint and Focal Length Optimization](https://mlsanity.com/paper/2503.01439/) (2025)

## Built on by (most cited first, 1 of 1)

- [FoveaVLA: Foveation Actions Replace the Wrist Camera in VLAs](https://mlsanity.com/paper/2610.20064/) (2026)

## Abstract (arXiv)

In embodied AI, visual perception should be active rather than passive: the system must decide where to look and at what scale to sense to acquire maximally informative data under pixel and spatial budget constraints. Existing vision models coupled with fixed RGB-D cameras fundamentally fail to reconcile wide-area coverage with fine-grained detail acquisition, severely limiting their efficacy in open-world robotic applications. We study the task of language-guided active visual perception: given a single RGB image and a natural language instruction, the agent must output pan, tilt, and zoom adjustments of a real PTZ (pan-tilt-zoom) camera to acquire the most informative view for the specified task. We propose EyeVLA, a unified framework that addresses this task by integrating visual perception, language understanding, and physical camera control within a single autoregressive vision-language-action model. EyeVLA introduces a semantically rich and efficient hierarchical action encoding that compactly tokenizes continuous camera adjustments and embeds them into the VLM vocabulary for joint multimodal reasoning. Through a data-efficient pipeline comprising pseudo-label generation, iterative IoU-controlled data refinement, and reinforcement learning with Group Relative Policy Optimization (GRPO), we transfer the open-world understanding of a pre-trained VLM to an embodied active perception policy using only 500 real-world samples. Evaluations on 50 diverse real-world scenes across five independent evaluation runs demonstrate that EyeVLA achieves an average task completion rate of 96%. Our work establishes a new paradigm for instruction-driven active visual information acquisition in multimodal embodied systems.