mlsanityData to 14 Oct 2026

Slow Perception: Let's Perceive Geometric Figures Step-by-step

arXiv 2412.20631 ↗First posted 30 Dec 2024cs.CVcited by 24

On the map: LLM reasoning → Spatial and embodied reasoning

Abstract (arXiv)

Recently, "visual o1" began to enter people's vision, with expectations that this slow-thinking design can solve visual reasoning tasks, especially geometric math problems. However, the reality is that current LVLMs (Large Vision Language Models) can hardly even accurately copy a geometric figure, let alone truly understand the complex inherent logic and spatial relationships within geometric shapes. We believe accurate copying (strong perception) is the first step to visual o1. Accordingly, we introduce the concept of "slow perception" (SP), which guides the model to gradually perceive basic point-line combinations, as our humans, reconstruct complex geometric structures progressively. There are two-fold stages in SP: a) perception decomposition. Perception is not instantaneous. In this stage, complex geometric figures are broken down into basic simple units to unify geometry representation. b) perception flow, which acknowledges that accurately tracing a line is not an easy task. This stage aims to avoid "long visual jumps" in regressing line segments by using a proposed "perceptual ruler" to trace each line stroke-by-stroke. Surprisingly, such a human-like perception manner enjoys an inference time scaling law -- the slower, the better. Researchers strive to speed up the model's perception in the past, but we slow it down again, allowing the model to read the image step-by-step and carefully.

Key facts. Cited by 24 papers in the map (data through 14 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 31 of its 37 references that are in the map. First posted on arXiv 30 Dec 2024.

Builds on

31 of 37 references.

References · 31

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  4. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  5. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  6. 2020End-to-End Object Detection with Transformerscited by 7,414
  7. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,045
  8. 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,852
  9. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  10. 2015Fast R-CNNcited by 2,933
  11. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,224
  12. 2013Rich feature hierarchies for accurate object detection and semantic segmentationcited by 2,162
19 more
  1. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,080
  2. 2019Towards VQA Models That Can Readcited by 1,718
  3. 2023MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contextscited by 1,614
  4. 2022ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoningcited by 1,479
  5. 2016YOLO9000: Better, Faster, Strongercited by 1,431
  6. 2020DocVQA: A Dataset for VQA on Document Imagescited by 1,317
  7. 2024How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suitescited by 1,050
  8. 2018CornerNet: Detecting Objects as Paired Keypointscited by 763
  9. 2021GeoQA: A Geometric Question Answering Benchmark Towards Multimodal Numerical Reasoningcited by 228
  10. 2024General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Modelcited by 180
  11. 2023mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document Understandingcited by 168
  12. 2024Vision language models are blind: Failing to translate detailed visual features into wordscited by 147
  13. 2023Vary: Scaling up the Vision Vocabulary for Large Vision-Language Modelscited by 84
  14. 2023ChatSpot: Bootstrapping Multimodal LLMs via Precise Referring Instruction Tuningcited by 65
  15. 2024Focus Anywhere for Fine-grained Multi-page Document Understandingcited by 62
  16. 2023Merlin:Empowering Multimodal LLMs with Foresight Mindscited by 46
  17. 2024Small Language Model Meets with Reinforced Vision Vocabularycited by 43
  18. 2024OneChart: Purify the Chart Structural Extraction via One Auxiliary Tokencited by 39
  19. 2024Historical Astronomical Diagrams Decomposition in Geometric Primitivescited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 18; 2026: 6.

24 papers.

Sort:

By area:

  1. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 384
  2. 2025From System 1 to System 2: A Survey of Reasoning Large Language Modelscited by 320
  3. 2025DeepSeek-OCR: Contexts Optical Compressioncited by 183
  4. 2025NoisyRollout: Reinforcing Visual Reasoning with Data Augmentationcited by 86
  5. 2026DeepSeek-OCR 2: Visual Causal Flowcited by 74
  6. 2025MathCanvas: Intrinsic Visual Chain-of-Thought for Multimodal Mathematical Reasoningcited by 35
  7. 2025AlphaOne: Reasoning Models Thinking Slow and Fast at Test Timecited by 29
  8. 2025Why Vision Language Models Struggle with Visual Arithmetic? Towards Enhanced Chart and Geometry Understandingcited by 24
  9. 2025Perception Before Reasoning: Two-Stage Reinforcement Learning for Visual Reasoning in Vision-Language Modelscited by 17
  10. 2025Perception in Reflectioncited by 13
  11. 2025MathFlow: Enhancing the Perceptual Flow of MLLMs for Visual Mathematical Problemscited by 10
  12. 2025PerPO: Perceptual Preference Optimization via Discriminative Rewardingcited by 9
12 more
  1. 2025VLMs have Tunnel Vision: Evaluating Nonlocal Visual Reasoning in Leading VLMscited by 7
  2. 2025CapGeo-Bench: Decoupling Visual Perception from Reasoning and Evaluating Geometric Understandingcited by 5
  3. 2026CogFlow: Bridging Perception and Reasoning through Knowledge Internalization for Visual Mathematical Problem Solvingcited by 5
  4. 2026OMIBench: Benchmarking Olympiad-Level Multi-Image Reasoning in Large Vision-Language Modelcited by 5
  5. 2025A Survey of Deep Learning for Geometry Problem Solvingcited by 4
  6. 2025MMGeoLM: Hard Negative Contrastive Learning for Fine-Grained Geometric Understanding in Large Multimodal Modelscited by 2
  7. 2025Thinking by Doing: Building Efficient World Model Reasoning in LLMs via Multi-turn Interactioncited by 2
  8. 2025InterFeedback: Unveiling Interactive Intelligence of Large Multimodal Models via Human Feedbackcited by 1
  9. 2026GeoFocus: Blending Efficient Global-to-Local Perception for Multimodal Geometry Problem-Solvingcited by 1
  10. 2025RoadBench: Benchmarking MLLMs on Fine-Grained Spatial Understanding and Reasoning under Urban Road Scenarioscited by 0
  11. 2026Concise Geometric Description as a Bridge: Unleashing the Potential of LLM for Plane Geometry Problem Solvingcited by 0
  12. 2026StemBind: When MLLMs Get Lost Between Rules and Instances in Abstract Visual Reasoningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Decoupling the components of geometric understanding in Vision Language Models
  2. 2024VisOnlyQA: Large Vision Language Models Still Struggle with Visual Perception of Geometric Information
  3. 2024GePBench: Evaluating Fundamental Geometric Perception for Multimodal Large Language Models
  4. 2026Reactivating Test-Time Scaling for Plane Geometry Problem Solving
  5. 2025Forgotten Polygons: Multimodal Large Language Models are Shape-Blind
  6. 2024Understanding Depth and Height Perception in Large Visual-Language Models
  7. 2026Beyond Generation and Accuracy: Diagnosing and Enhancing Visual Chain-of-Thought for Geometry Problem Solving
  8. 2026From Symbolic Perception to Logical Deduction: A Framework for Guiding Language Models in Geometric Reasoning