Human Cognitive Benchmarks Reveal Foundational Visual Gaps in MLLMs

• URL: http://arxiv.org/abs/2502.16435v2
• Date: Thu, 07 Aug 2025 15:39:27 GMT
• Title: Human Cognitive Benchmarks Reveal Foundational Visual Gaps in MLLMs
• Authors: Jen-Tse Huang, Dasen Dai, Jen-Yuan Huang, Youliang Yuan, Xiaoyuan Liu, Wenxuan Wang, Wenxiang Jiao, Pinjia He, Zhaopeng Tu, Haodong Duan,
• Abstract summary: VisFactor is a benchmark that digitizes 20 vision-centric subtests from a well-established cognitive psychology assessment.<n>We evaluate 20 frontier Multimodal Large Language Models (MLLMs) from GPT, Gemini, Claude, LLaMA, Qwen, and SEED families.<n>The best-performing model achieves a score of only 25.19 out of 100, with consistent failures on tasks such as mental rotation, spatial relation inference, and figure-ground discrimination.
• Score: 65.93003087656754
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Despite significant progress on popular multimodal benchmarks, state-of-the-art Multimodal Large Language Models (MLLMs) continue to struggle with basic visual reasoning tasks that are trivially solved by humans, such as recognizing spatial relationships. To systematically investigate this gap, we introduce VisFactor, a benchmark that digitizes 20 vision-centric subtests from a well-established cognitive psychology assessment. These subtests span four core domains of human visual cognition: (1) Visualization and Spatial Processing, (2) Perceptual and Closure, (3) Memory, and (4) Reasoning. We evaluate 20 frontier MLLMs from GPT, Gemini, Claude, LLaMA, Qwen, and SEED families. The best-performing model achieves a score of only 25.19 out of 100, with consistent failures on tasks such as mental rotation, spatial relation inference, and figure-ground discrimination, regardless of model size or prompting strategy. These findings suggest that current MLLM performance gains on high-level benchmarks do not reflect human-like low-level visual cognition, challenging the assumption that large-scale pretraining naturally induces gestalt-like perceptual capabilities. The dataset and evaluation toolkit are publicly available at: https://github.com/CUHK-ARISE/VisFactor.

Related papers

• Beyond the Visible: Benchmarking Occlusion Perception in Multimodal Large Language Models [17.922450921582794]
  Occlusion perception is a critical foundation for human-level spatial understanding.<n>We introduce O-Bench, the first visual question answering (VQA) benchmark specifically designed for occlusion perception.
  arXiv  Detail & Related papers  (2025-08-06T03:39:21Z)
• SpatialViz-Bench: Automatically Generated Spatial Visualization Reasoning Tasks for MLLMs [43.82781630267406]
  SpatialViz-Bench is a comprehensive benchmark for spatial visualization with 12 tasks across 4 sub-abilities, comprising 1,180 automatically generated problems.<n>Our evaluation of 33 state-of-the-art MLLMs reveals wide performance variations and uncovers counter-intuitive findings.
  arXiv  Detail & Related papers  (2025-07-10T10:27:20Z)
• Jigsaw-Puzzles: From Seeing to Understanding to Reasoning in Vision-Language Models [6.569837864665502]
  We introduce Jigsaw-Puzzles, a novel benchmark consisting of 1,100 carefully curated real-world images with high spatial complexity.<n>Based on this dataset, we design five tasks to rigorously evaluate vision-language models' spatial perception, structural understanding, and reasoning capabilities.<n>The results show that even the strongest model, Gemini-2.5-Pro, achieves only 77.14% overall accuracy and performs particularly poorly on the Order Generation task.
  arXiv  Detail & Related papers  (2025-05-27T05:17:41Z)
• VisuLogic: A Benchmark for Evaluating Visual Reasoning in Multi-modal Large Language Models [121.03333569013148]
  We introduce VisuLogic: a benchmark of 1,000 human-verified problems across six categories. These types of questions can be evaluated to assess the visual reasoning capabilities of MLLMs from multiple perspectives. Most models score below 30% accuracy-only slightly above the 25% random baseline and far below the 51.4% achieved by humans.
  arXiv  Detail & Related papers  (2025-04-21T17:59:53Z)
• Benchmarking Large Vision-Language Models on Fine-Grained Image Tasks: A Comprehensive Evaluation [53.84282335629258]
  We introduce a comprehensive fine-grained evaluation benchmark, i.e., FG-BMK, comprising 3.49 million questions and 3.32 million images. Our evaluation systematically examines LVLMs from both human-oriented and machine-oriented perspectives. We uncover key findings regarding the influence of training paradigms, modality alignment, perturbation susceptibility, and fine-grained category reasoning on task performance.
  arXiv  Detail & Related papers  (2025-04-21T09:30:41Z)
• Multimodal LLM Augmented Reasoning for Interpretable Visual Perception Analysis [19.032828729570458]
  We use established principles and explanations from psychology and cognitive science related to complexity in human visual perception. Our study aims to benchmark MLLMs across various explainability principles relevant to visual perception.
  arXiv  Detail & Related papers  (2025-04-16T22:14:27Z)
• Expanding the Boundaries of Vision Prior Knowledge in Multi-modal Large Language Models [53.13731845500678]
  We introduce a novel metric, $Rank_e$, to quantify the effect of vision encoder's prior knowledge on MLLM performance. We propose VisPRE, a two-stage training framework that explicitly incorporates prior knowledge at the vision encoder level. Experimental results demonstrate that augmenting vision encoder's prior knowledge substantially boosts the visual understanding capabilities of MLLMs.
  arXiv  Detail & Related papers  (2025-03-23T11:33:09Z)
• Grounded Chain-of-Thought for Multimodal Large Language Models [66.04061083611863]
  We propose a new learning task for multimodal large language models (MLLMs) called Grounded Chain-of-Thought (GCoT) GCoT is keen to helping MLLMs to recognize and ground the relevant visual cues step by step, thereby predicting the correct answer with grounding coordinates as the intuitive basis. To facilitate this task, we also carefully design and construct a dataset called multimodal grounded chain-of-thought (MM-GCoT) consisting of 24,022 GCoT examples for 5,033 images.
  arXiv  Detail & Related papers  (2025-03-17T04:07:47Z)
• DeepPerception: Advancing R1-like Cognitive Visual Perception in MLLMs for Knowledge-Intensive Visual Grounding [61.26026947423187]
  Human experts excel at fine-grained visual discrimination by leveraging domain knowledge to refine perceptual features.<n>Current Multimodal Large Language Models (MLLMs) struggle to integrate reasoning into visual perception.<n>We propose DeepPerception, an MLLM enhanced with cognitive visual perception capabilities.
  arXiv  Detail & Related papers  (2025-03-17T04:06:34Z)
• VOILA: Evaluation of MLLMs For Perceptual Understanding and Analogical Reasoning [63.0285363282581]
  Multimodal Large Language Models (MLLMs) have become a powerful tool for integrating visual and textual information.<n>We introduce VOILA, a benchmark designed to evaluate MLLMs' perceptual understanding and abstract relational reasoning.<n>We reveal that current MLLMs struggle to comprehend inter-image relationships and exhibit limited capabilities in high-level relational reasoning.
  arXiv  Detail & Related papers  (2025-02-25T23:36:19Z)
• Human Re-ID Meets LVLMs: What can we expect? [14.370360290704197]
  We compare the performance of the leading large vision-language models in the human re-identification task.<n>Our results confirm the strengths of LVLMs, but also their severe limitations that often lead to catastrophic answers.
  arXiv  Detail & Related papers  (2025-01-30T19:00:40Z)
• GePBench: Evaluating Fundamental Geometric Perception for Multimodal Large Language Models [34.647839550142834]
  We introduce GePBench, a novel benchmark designed to assess the geometric perception abilities of MLLMs.<n>Our evaluations reveal that current state-of-the-art MLLMs exhibit significant deficiencies in geometric perception tasks.<n>We show that models trained with GePBench data demonstrate substantial improvements on a wide range of benchmark tasks.
  arXiv  Detail & Related papers  (2024-12-30T16:01:43Z)
• Understanding the Role of LLMs in Multimodal Evaluation Benchmarks [77.59035801244278]
  This paper investigates the role of the Large Language Model (LLM) backbone in Multimodal Large Language Models (MLLMs) evaluation. Our study encompasses four diverse MLLM benchmarks and eight state-of-the-art MLLMs. Key findings reveal that some benchmarks allow high performance even without visual inputs and up to 50% of error rates can be attributed to insufficient world knowledge in the LLM backbone.
  arXiv  Detail & Related papers  (2024-10-16T07:49:13Z)
• Reasoning Paths with Reference Objects Elicit Quantitative Spatial Reasoning in Large Vision-Language Models [61.899791071654654]
  We introduce a benchmark, Q-Spatial Bench, with 271 questions across five categories designed for quantitative spatial reasoning. We investigate the performance of state-of-the-art vision-language models (VLMs) on this task. We develop a zero-shot prompting technique, SpatialPrompt, that encourages VLMs to answer quantitative spatial questions using reference objects as visual cues.
  arXiv  Detail & Related papers  (2024-09-15T16:45:42Z)
• Visualization Literacy of Multimodal Large Language Models: A Comparative Study [12.367399155606162]
  multimodal large language models (MLLMs) combine the inherent power of large language models (LLMs) with the renewed capabilities to reason about the multimodal context. Many recent works in visualization have demonstrated MLLMs' capability to understand and interpret visualization results and explain the content of the visualization to users in natural language. In this work, we aim to fill the gap by utilizing the concept of visualization literacy to evaluate MLLMs.
  arXiv  Detail & Related papers  (2024-06-24T17:52:16Z)
• What is the Visual Cognition Gap between Humans and Multimodal LLMs? [22.99627171182423]
  Multimodal Large Language Models (MLLMs) have shown great promise in language-guided tasks such as recognition, segmentation, and object detection. One such challenge is abstract visual reasoning (AVR) -- the cognitive ability to discern relationships among patterns in a set of images and extrapolate to predict subsequent patterns. We propose new dataset MaRs-VQA and a new benchmark VCog-Bench to evaluate the zero-shot capability of MLLMs.
  arXiv  Detail & Related papers  (2024-06-14T22:02:21Z)
• Hallucination of Multimodal Large Language Models: A Survey [40.73148186369018]
  multimodal large language models (MLLMs) have demonstrated significant advancements and remarkable abilities in multimodal tasks. Despite these promising developments, MLLMs often generate outputs that are inconsistent with the visual content. This survey aims to deepen the understanding of hallucinations in MLLMs and inspire further advancements in the field.
  arXiv  Detail & Related papers  (2024-04-29T17:59:41Z)
• VALOR-EVAL: Holistic Coverage and Faithfulness Evaluation of Large Vision-Language Models [57.43276586087863]
  Large Vision-Language Models (LVLMs) suffer from hallucination issues, wherein the models generate plausible-sounding but factually incorrect outputs. Existing benchmarks are often limited in scope, focusing mainly on object hallucinations. We introduce a multi-dimensional benchmark covering objects, attributes, and relations, with challenging images selected based on associative biases.
  arXiv  Detail & Related papers  (2024-04-22T04:49:22Z)
• NPHardEval4V: A Dynamic Reasoning Benchmark of Multimodal Large Language Models [34.91372939329467]
  We introduce a benchmark, NPHardEval4V, to evaluate the pure reasoning abilities of MLLMs. Our findings reveal significant discrepancies in reasoning abilities across different models. We also investigate the impact of different prompting styles, including visual, text, and combined visual and text prompts, on the reasoning abilities of MLLMs.
  arXiv  Detail & Related papers  (2024-03-04T07:10:31Z)
• Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMs [50.77984109941538]
  Our research reveals that the visual capabilities in recent multimodal LLMs still exhibit systematic shortcomings. We identify ''CLIP-blind pairs'' - images that CLIP perceives as similar despite their clear visual differences. We evaluate various CLIP-based vision-and-language models and found a notable correlation between visual patterns that challenge CLIP models and those problematic for multimodal LLMs.
  arXiv  Detail & Related papers  (2024-01-11T18:58:36Z)
• Behind the Magic, MERLIM: Multi-modal Evaluation Benchmark for Large Image-Language Models [50.653838482083614]
  This paper introduces a scalable test-bed to assess the capabilities of IT-LVLMs on fundamental computer vision tasks. MERLIM contains over 300K image-question pairs and has a strong focus on detecting cross-modal "hallucination" events in IT-LVLMs.
  arXiv  Detail & Related papers  (2023-12-03T16:39:36Z)
• From CLIP to DINO: Visual Encoders Shout in Multi-modal Large Language Models [36.41816380074965]
  We investigate the effectiveness of different vision encoders within Large Language Models (MLLMs) Our findings reveal that the shallow layer features of CLIP offer particular advantages for fine-grained tasks such as grounding and region understanding. We propose a simple yet effective feature merging strategy, named COMM, that integrates CLIP and DINO with Multi-level features Merging.
  arXiv  Detail & Related papers  (2023-10-13T02:41:55Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.