Search papers, labs, and topics across Lattice.
The paper introduces Vision Banana, a generalist vision model built by instruction-tuning Nano Banana Pro (NBP) on a mixture of its original training data and vision task data, reframing perception as image generation. This approach achieves state-of-the-art results on diverse 2D and 3D vision tasks, rivaling or surpassing zero-shot domain-specific models like SAM and Depth Anything. The results demonstrate that image generation pretraining can serve as a powerful generalist vision learner, similar to how language model pretraining enables emergent capabilities.
Image generators aren't just for making pretty pictures; they're secretly state-of-the-art vision learners, rivaling specialized models in tasks from segmentation to depth estimation.
Recent works show that image and video generators exhibit zero-shot visual understanding behaviors, in a way reminiscent of how LLMs develop emergent capabilities of language understanding and reasoning from generative pretraining. While it has long been conjectured that the ability to create visual content implies an ability to understand it, there has been limited evidence that generative vision models have developed strong understanding capabilities. In this work, we demonstrate that image generation training serves a role similar to LLM pretraining, and lets models learn powerful and general visual representations that enable SOTA performance on various vision tasks. We introduce Vision Banana, a generalist model built by instruction-tuning Nano Banana Pro (NBP) on a mixture of its original training data alongside a small amount of vision task data. By parameterizing the output space of vision tasks as RGB images, we seamlessly reframe perception as image generation. Our generalist model, Vision Banana, achieves SOTA results on a variety of vision tasks involving both 2D and 3D understanding, beating or rivaling zero-shot domain-specialists, including Segment Anything Model 3 on segmentation tasks, and the Depth Anything series on metric depth estimation. We show that these results can be achieved with lightweight instruction-tuning without sacrificing the base model's image generation capabilities. The superior results suggest that image generation pretraining is a generalist vision learner. It also shows that image generation serves as a unified and universal interface for vision tasks, similar to text generation's role in language understanding and reasoning. We could be witnessing a major paradigm shift for computer vision, where generative vision pretraining takes a central role in building Foundational Vision Models for both generation and understanding.
