Paper page - WeVisOne-VL: Unified Visual Perception from Objects to Pixels

Source: https://huggingface.co/papers/2610.31329

Abstract

WeVisOne-VL unifies object-level and pixel-level visual perception via structured token prediction, supporting detection, localization, counting, OCR, segmentation, and matting with a lightweight pixel decoder and multi-stage training.

Existingvision-language models(VLMs) remain unreliable for fine-grained visual perception. In this report, we present WeVisOne-VL (WeChat Vision All-in-One), a powerful VLM that unifies object-level andpixel-level perceptionthroughstructured next-token prediction. It supports multiple visual tasks including object detection, referring localization, counting, OCR, segmentation, and matting, enabling localization at any user-specific granularity, from points and boxes to masks and alpha mattes. Architecturally, we integrate a lightweightpixel decoderinto the VLM via a special<OBJ> tokento generate instance-level pixel outputs. We train the model with a three-stagecurriculum SFTfollowed bymixed-task RL. Across over 30 benchmarks, WeVisOne-VL demonstrates leading performance on both object-level and pixel-level tasks, while maintaining robust rejection forobject-absent queries. By releasing WeVisOne-VL to the open-source community, we hope to advance VLMs from broad visual understanding toward truly reliable perception systems for real-world deployment.

View arXiv pageView PDFProject pageGitHub310Add to collection

Models citing this paper2

#### tencent/WeVisOne-VL-4B Image-Text-to-Text• 5B• Updatedabout 21 hours ago • 3.79k • 67 #### tencent/WeVisOne-VL-2B Image-Text-to-Text• 2B• Updatedabout 21 hours ago • 3.09k • 51

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2610.31329 in a dataset README.md to link it from this page.

Spaces citing this paper4

Collections including this paper7

Browse 7 collections that include this paper