paper-with-me

홈 › Papers

Improved Visual-Spatial Reasoning via R1-Zero-Like Training

2025-04-01 · Zhenyi Liao, Qingsong Xie, Yanhao Zhang, Zijian Kong, Haonan Lu, Zhenyu Yang, Zhijie Deng

Increasing attention has been placed on improving the reasoning capacities of multi-modal large language models (MLLMs). As the cornerstone for AI agents that function in the physical realm, video-based visual-spatial intelligence (VSI) emerges as one of the most pivotal reasoning capabilities of MLLMs. This work conducts a first, in-depth study on improving the visual-spatial reasoning of MLLMs via R1-Zero-like training. Technically, we first identify that the visual-spatial reasoning capacities of small- to medium-sized Qwen2-VL models cannot be activated via Chain of Thought (CoT) prompts. We then incorporate GRPO training for improved visual-spatial reasoning, using the carefully curated VSI-100k dataset, following DeepSeek-R1-Zero. During the investigation, we identify the necessity to keep the KL penalty (even with a small value) in GRPO. With just 120 GPU hours, our vsGRPO-2B model, fine-tuned from Qwen2-VL-2B, can outperform the base model by 12.1% and surpass GPT-4o. Moreover, our vsGRPO-7B model, fine-tuned from Qwen2-VL-7B, achieves performance comparable to that of the best open-source model LLaVA-NeXT-Video-72B. Additionally, we compare vsGRPO to supervised fine-tuning and direct preference optimization baselines and observe strong performance superiority. The code and dataset will be available soon.

📄 PDF Abstract BibTeX arXiv:2504.00883

Code (1)

zhijie-group/r1-zero-vsi

Tasks

GPUSpatial Reasoning

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음
BASE 설명 없음

Similar Papers 제목 키워드 기반

I Know About "Up"! Enhancing Spatial Reasoning in Visual Language Models Through 3D Reconstruction

2024-07-19 · Zaiqiao Meng, Hao Zhou, Yifang Chen

Visual Language Models (VLMs) are essential for various tasks, particularly visual reasoning tasks, due to their robust multi-modal information integration, visual reasoning capabilities, and contextual awareness. Howeve…

3D ReconstructionSpatial ReasoningVisual Reasoning

Can Vision Language Models Learn from Visual Demonstrations of Ambiguous Spatial Reasoning?

2024-09-25 · Bowen Zhao, Leo Parker Dirac, Paulina Varshavskaya

Large vision-language models (VLMs) have become state-of-the-art for many computer vision tasks, with in-context learning (ICL) as a popular adaptation strategy for new ones. But can VLMs learn novel concepts purely from…

In-Context LearningNovel ConceptsSpatial Reasoning

SpatialPrompting: Keyframe-driven Zero-Shot Spatial Reasoning with Off-the-Shelf Multimodal Large Language Models

2025-05-08 · Shun Taguchi, Hideki Deguchi, Takumi Hamazaki, Hiroyuki Sakai

This study introduces SpatialPrompting, a novel framework that harnesses the emergent reasoning capabilities of off-the-shelf multimodal large language models to achieve zero-shot spatial reasoning in three-dimensional (…

Spatial Reasoning

TopV-Nav: Unlocking the Top-View Spatial Reasoning Potential of MLLM for Zero-shot Object Navigation

2024-11-25 · Linqing Zhong, Chen Gao, Zihan Ding, Yue Liao 외

The Zero-Shot Object Navigation (ZSON) task requires embodied agents to find a previously unseen object by navigating in unfamiliar environments. Such a goal-oriented exploration heavily relies on the ability to perceive…

Spatial Reasoning

3D-Layout-R1: Structured Reasoning for Language-Instructed Spatial Editing

2026-03-23 · Haoyu Zhen, Xiaolong Li, Yilin Zhao, Han Zhang 외 arxiv

Large Language Models (LLMs) and Vision Language Models (VLMs) have shown impressive reasoning abilities, yet they struggle with spatial understanding and layout consistency when performing fine-grained visual editing. W…