Blind Spot Navigation: Evolutionary Discovery of Sensitive Semantic Concepts for LVLMs

• URL: http://arxiv.org/abs/2505.15265v2
• Date: Fri, 25 Jul 2025 09:03:08 GMT
• Title: Blind Spot Navigation: Evolutionary Discovery of Sensitive Semantic Concepts for LVLMs
• Authors: Zihao Pan, Yu Tong, Weibin Wu, Jingyi Wang, Lifeng Chen, Zhe Zhao, Jiajia Wei, Yitong Qiao, Zibin Zheng,
• Abstract summary: Recent research suggests that models may be particularly sensitive to certain semantics in visual inputs, making them prone to errors.<n>Inspired by this, in this paper we conducted the first exploration on large vision-language models (LVLMs)<n>We found that LVLMs indeed are susceptible to hallucinations and various errors when facing specific semantic concepts in images.
• Score: 24.76767896607915
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Adversarial attacks aim to generate malicious inputs that mislead deep models, but beyond causing model failure, they cannot provide certain interpretable information such as ``\textit{What content in inputs make models more likely to fail?}'' However, this information is crucial for researchers to specifically improve model robustness. Recent research suggests that models may be particularly sensitive to certain semantics in visual inputs (such as ``wet,'' ``foggy''), making them prone to errors. Inspired by this, in this paper we conducted the first exploration on large vision-language models (LVLMs) and found that LVLMs indeed are susceptible to hallucinations and various errors when facing specific semantic concepts in images. To efficiently search for these sensitive concepts, we integrated large language models (LLMs) and text-to-image (T2I) models to propose a novel semantic evolution framework. Randomly initialized semantic concepts undergo LLM-based crossover and mutation operations to form image descriptions, which are then converted by T2I models into visual inputs for LVLMs. The task-specific performance of LVLMs on each input is quantified as fitness scores for the involved semantics and serves as reward signals to further guide LLMs in exploring concepts that induce LVLMs. Extensive experiments on seven mainstream LVLMs and two multimodal tasks demonstrate the effectiveness of our method. Additionally, we provide interesting findings about the sensitive semantics of LVLMs, aiming to inspire further in-depth research.

Related papers

• SECOND: Mitigating Perceptual Hallucination in Vision-Language Models via Selective and Contrastive Decoding [5.976839106353883]
  SECOND: Selective and Contrastive Decoding is a novel approach that enables Vision-Language Models to leverage multi-scale visual information with an object-centric manner.<n> SECOND significantly reduces perceptual hallucinations and outperforms a wide range of benchmarks.
  arXiv  Detail & Related papers  (2025-06-10T02:55:38Z)
• E2LVLM:Evidence-Enhanced Large Vision-Language Model for Multimodal Out-of-Context Misinformation Detection [7.1939657372410375]
  We present E2LVLM, a novel evidence-enhanced large vision-language model by adapting textual evidence in two levels.<n>To address the scarcity of news domain datasets with both judgment and explanation, we generate a novel OOC multimodal instruction-following dataset.<n>A multitude of experiments demonstrate that E2LVLM achieves superior performance than state-of-the-art methods.
  arXiv  Detail & Related papers  (2025-02-12T04:25:14Z)
• VladVA: Discriminative Fine-tuning of LVLMs [67.14293827774827]
  Contrastively-trained Vision-Language Models (VLMs) like CLIP have become the de facto approach for discriminative vision-language representation learning.<n>We propose to combine "the best of both worlds": a new training approach for discriminative fine-tuning of LVLMs.
  arXiv  Detail & Related papers  (2024-12-05T17:54:27Z)
• When Large Vision-Language Models Meet Person Re-Identification [44.604485649167216]
  We propose LVLM-ReID, a novel framework that harnesses the strengths of LVLMs to promote ReID.<n>Our framework integrates the semantic understanding and generation capabilities of LVLMs into end-to-end ReID training.<n>Our method achieves competitive results on multiple benchmarks without additional image-text annotations.
  arXiv  Detail & Related papers  (2024-11-27T07:45:25Z)
• Looking Beyond Text: Reducing Language bias in Large Vision-Language Models via Multimodal Dual-Attention and Soft-Image Guidance [67.26434607115392]
  Large vision-language models (LVLMs) have achieved impressive results in various vision-language tasks. LVLMs suffer from hallucinations caused by language bias, leading to diminished focus on images and ineffective visual comprehension. We propose LACING to address the language bias of LVLMs with muLtimodal duAl-attention meChanIsm (MDA) aNd soft-image Guidance (IFG)
  arXiv  Detail & Related papers  (2024-11-21T16:33:30Z)
• Large Vision-Language Models as Emotion Recognizers in Context Awareness [14.85890824622433]
  Context-aware emotion recognition (CAER) is a complex and significant task that requires perceiving emotions from various contextual cues. Previous approaches primarily focus on designing sophisticated architectures to extract emotional cues from images. This paper systematically explore the potential of leveraging Large Vision-Language Models (LVLMs) to empower the CAER task.
  arXiv  Detail & Related papers  (2024-07-16T01:28:06Z)
• From Redundancy to Relevance: Information Flow in LVLMs Across Reasoning Tasks [33.476693301050275]
  We conduct experiments with truncation strategies across various LVLMs for visual question answering and image captioning tasks. By exploring the information flow from the perspective of visual representation contribution, we observe that it tends to converge in shallow layers but diversify in deeper layers.
  arXiv  Detail & Related papers  (2024-06-04T13:52:54Z)
• Chain-of-Spot: Interactive Reasoning Improves Large Vision-Language Models [81.71651422951074]
  Chain-of-Spot (CoS) method is a novel approach that enhances feature extraction by focusing on key regions of interest. This technique allows LVLMs to access more detailed visual information without altering the original image resolution. Our empirical findings demonstrate a significant improvement in LVLMs' ability to understand and reason about visual content.
  arXiv  Detail & Related papers  (2024-03-19T17:59:52Z)
• Finer: Investigating and Enhancing Fine-Grained Visual Concept Recognition in Large Vision Language Models [57.95366341738857]
  In-depth analyses show that instruction-tuned LVLMs exhibit modality gap, showing discrepancy when given textual and visual inputs that correspond to the same concept.<n>We propose a multiple attribute-centric evaluation benchmark, Finer, to evaluate LVLMs' fine-grained visual comprehension ability and provide significantly improved explainability.
  arXiv  Detail & Related papers  (2024-02-26T05:43:51Z)
• Mitigating Object Hallucination in Large Vision-Language Models via Image-Grounded Guidance [51.30560006045442]
  Image-gRounded guIdaNcE (MARINE) is a framework that is both training-free and API-free.<n>MARINE effectively and efficiently reduces object hallucinations during inference by introducing image-grounded guidance to LVLMs.<n>Our framework's flexibility further allows for the integration of multiple vision models, enabling more reliable and robust object-level guidance.
  arXiv  Detail & Related papers  (2024-02-13T18:59:05Z)
• LION : Empowering Multimodal Large Language Model with Dual-Level Visual Knowledge [58.82222646803248]
  Multimodal Large Language Models (MLLMs) have endowed LLMs with the ability to perceive and understand multi-modal signals. Most of the existing MLLMs mainly adopt vision encoders pretrained on coarsely aligned image-text pairs, leading to insufficient extraction and reasoning of visual knowledge. We propose a dual-Level vIsual knedgeOwl eNhanced Multimodal Large Language Model (LION), which empowers the MLLM by injecting visual knowledge in two levels.
  arXiv  Detail & Related papers  (2023-11-20T15:56:44Z)
• HiLM-D: Enhancing MLLMs with Multi-Scale High-Resolution Details for Autonomous Driving [44.06475712570428]
  HiLM-D is a resource-efficient framework that enhances visual information processing in MLLMs for ROLISP.<n>Our method is motivated by the fact that the primary variations in autonomous driving scenarios are the motion trajectories.<n>Experiments show HiLM-D's significant improvements over current MLLMs, with a 3.7% in BLEU-4 for captioning and 8.7% in mIoU for detection.
  arXiv  Detail & Related papers  (2023-09-11T01:24:13Z)

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.