This work introduces Scene Graph Thinking (SaGe), a novel paradigm that enables fine-grained and structured visual reasoning through explicit scene-graph representations and proposes node-as-proxy graph rewards to consolidate efficient graph exploration.
Abstract
Multimodal Large Language Models (MLLMs) have demonstrated strong perception and reasoning capabilities. However, most existing models focus on isolated objects and neglect structured relationships for efficient target navigation, limiting their performance on visually intensive tasks. To address this challenge, we introduce Scene Graph Thinking (SaGe), a novel paradigm that enables fine-grained and structured visual reasoning through explicit scene-graph representations. Specifically, we first introduce an automated data engine that converts flat image-text corpora into structured scene graphs, where hierarchical entities constitute the nodes and diverse visual relations define the edges. Building upon this, we construct 120K high-quality training data by sampling reasoning traces from scene graphs. Then, two-stage graph-aligned post-training paradigms are introduced, where supervised fine-tuning internalizes MLLMs with structured reasoning, and subsequent reinforcement fine-tuning proposes node-as-proxy graph rewards to consolidate efficient graph exploration. With curated data and graph-aligned training, our approach achieves significant improvements across eight multimodal benchmarks, demonstrating strong effectiveness on fine-grained perception and reasoning tasks. Code is available at https://github.com/zwyang6/SaGe.
GraphVerse is introduced, a unified benchmark that jointly evaluates perception, visual reasoning, and text-based graph reasoning in MLLMs under both single-image and paired-image settings and proposes VGR-Score, a process-sensitive metric that evaluates reasoning quality beyond final-answer accuracy.
Yuanfu Sun, Yuanhang Ren, Kang Li et al.· 0 citations
Multimodal Large Language Models (MLLMs) capable of thinking with images often rely on external tools for fine-grained perception. However, this reliance introduces significant inference latency and fails to effectively resolve the spatial-structural gap-a fundamental challenge in text-dense and structurally relational visuals (e.g., charts and visual tables) where strict relative spatial arrangements bind textual elements. Without external tools, standard MLLMs struggle with such fine-grained visual reasoning tasks. To address these issues, we propose Think with Structured Grounding (TwSG), a novel fine-grained image perception framework designed to internalize complex images's tool-use capabilities within the model. TwSG distills the benefits of multi-step reasoning and micro-cropping into a single efficient forward pass during inference. Specifically, we use an MLLM to identify key regions guided by ground-truth answers, and then prompt a teacher model to generate high-quality visual question-answering (VQA) data. These fine-grained, region-based supervisory signals are subsequently distilled back into the full-image representation. Our training pipeline consists of two stages: (1) a cold-start supervised fine-tuning (SFT) phase using multi-turn data with focused area descriptions to foster complex reasoning and error recovery; and (2) a reinforcement fine-tuning (RFT) phase driven by a novel process reward mechanism, TL-GRPO, which encourages strategic reasoning. Extensive experiments across various MLLM architectures demonstrate that TwSG reduces inference latency while substantially improving accuracy and robustness, endowing models with native fine-grained region description and flexible reasoning capabilities.
Changjiang Jiang, Qiannian Zhao, Lei Xin et al.· 0 citations
Vision-Language Models (VLMs) have achieved remarkable progress in aligning visual and textual information; however, their inference processes remain largely implicit, end-to-end, and weakly structured. As a result, even state-of-the-art models often struggle with logical consistency, spatial reasoning, multi-entity binding, and robustness to occlusion or viewpoint changes, limiting their reliability in scientific, industrial, and safety-critical applications. To address these limitations, we propose Structure-Aware Visual Reasoning (SAVR), a conceptual framework that augments VLMs with explicit representation and constraint-based reasoning components. SAVR decomposes visual inference into three interpretable stages: (i) an entity–attribute–relation (EAR) representation, (ii) constraint-aware reasoning over structured representations, and (iii) integration of structured predictions with VLM-generated outputs. This formulation treats visual reasoning as an explicit process that enforces spatial, logical, and physical constraints over grounded entities, thereby improving consistency, transparency, and controllability. Furthermore, SAVR unifies several previously fragmented research directions-including scene graphs, object-centric learning, neuro-symbolic reasoning, and spatially grounded VLMs-into a coherent architectural perspective. Through this synthesis, we clarify how explicit structural representations can systematically mitigate common failure modes of end-to-end VLMs. While this work is conceptual and does not include empirical evaluation, it provides rigorous problem formulation, design rationale, and comparative analysis that establish a foundation for future structure-aware multimodal reasoning systems.
Scene graphs provide structured visual knowledge that can enhance multimodal large language models (MLLMs) for tasks like visual question answering and image captioning. However, existing approaches inject scene graphs as deterministic, hard prompts, ignoring the inherent uncertainty in visual perception, leading to overconfident and sometimes hallucinated outputs. We propose Probabilistic Scene Graph Prompting (PSGP), a framework that models scene graph generation as a distribution over plausible graphs and encodes this uncertainty into soft, continuous prompt tokens that condition the MLLM. By propagating perceptual uncertainty from detection to language generation, PSGP produces more accurate, faithful, and better-calibrated responses, especially in ambiguous visual scenarios. Experiments on GQA, Visual Spatial Reasoning, and a new Ambiguous-GQA benchmark show that PSGP outperforms strong baselines: including LLaVA-1.5, BLIP-2, and SG-LLaVA, in accuracy, faithfulness, and calibration, while maintaining computational efficiency. Our work establishes a principled pathway toward uncertainty-aware, structured multimodal intelligence.