Paper 2602.11858
Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception
- Published
- Feb 2026
- Research lab
- Independent
- Citations
- 23
- GitHub
- 184 stars
01 In brief
Summary
The paper introduces Region-to-Image Distillation (R2I), a method that internalizes the benefits of inference-time zooming into a single forward pass of a multimodal large language model (MLLM).
R2I synthesizes fine-grained VQA data by zooming into micro-cropped regions, using strong teacher models to generate high-consensus question-answer pairs, and then distilling this region-grounded supervision back to the full image with bounding-box overlays.
The authors also present ZoomBench, a benchmark of 845 VQA samples across six fine-grained perceptual dimensions, with a dual-view protocol to quantify the global-regional 'zooming gap'.
Experiments show that models trained with R2I (ZwZ) achieve leading performance on fine-grained perception benchmarks, outperform agentic 'Thinking-with-Images' models with lower latency, and improve general multimodal cognition.
The paper discusses when tool-based image operations are necessary versus when their benefits can be distilled into a single pass.
02 From the paper
Abstract
Multimodal Large Language Models (MLLMs) excel at broad visual understanding but still struggle with fine-grained perception, where decisive evidence is small and easily overwhelmed by global context. Recent "Thinking-with-Images" methods alleviate this by iteratively zooming in and out regions of interest during inference, but incur high latency due to repeated tool calls and visual re-encoding. To address this, we propose Region-to-Image Distillation, which transforms zooming from an inference-time tool into a training-time primitive, thereby internalizing the benefits of agentic zooming into a single forward pass of an MLLM. In particular, we first zoom in to micro-cropped regions to let strong teacher models generate high-quality VQA data, and then distill this region-grounded supervision back to the full image. After training on such data, the smaller student model improves "single-glance" fine-grained perception without tool use. To rigorously evaluate this capability, we further present ZoomBench, a hybrid-annotated benchmark of 845 VQA data spanning six fine-grained perceptual dimensions, together with a dual-view protocol that quantifies the global--regional "zooming gap". Experiments show that our models achieve leading performance across multiple fine-grained perception benchmarks, and also improve general multimodal cognition on benchmarks such as visual reasoning and GUI agents. We further discuss when "Thinking-with-Images" is necessary versus when its gains can be distilled into a single forward pass. Our code is available at https://github.com/inclusionAI/Zooming-without-Zooming.