English(EN) Robust Grounding with MLLMs against Occlusion and Small Objects via Language-guided Semantic Cues

MLLM利用语言引导的语义线索提升在拥挤场景下的目标定位能力

作者 PulseAugur 编辑部 · [2 个来源] · 2026-04-27 04:42

研究人员开发了一种新方法，以提高多模态大语言模型（MLLM）在拥挤场景等具有挑战性的视觉场景中的鲁棒性。该方法利用语言引导的语义线索（LGSC）来克服遮挡和小目标造成的性能下降问题。通过从MLLM的视觉管道中提取语义线索并用文本嵌入进行引导，该方法创建了语言语义先验，以精炼目标语义并提高定位准确性。 AI

影响增强了MLLM在复杂视觉环境中的鲁棒性，有望改进需要精确目标识别和定位的应用。

排序理由这是一篇详细介绍改进MLLM在特定任务上性能的新方法的学术论文。

在 arXiv cs.CV 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.CV TIER_1 English(EN) · Beomchan Park, Seongho Kim, Hyunjun Kim, Sungjune Park, Yong Man Ro · 2026-04-28 04:00

Robust Grounding with MLLMs against Occlusion and Small Objects via Language-guided Semantic Cues

arXiv:2604.24036v1 Announce Type: new Abstract: While Multimodal Large Language Models (MLLMs) have enhanced grounding capabilities in general scenes, their robustness in crowded scenes remains underexplored. Crowded scenes entail visual challenges (i.e., occlusion and small obje…
arXiv cs.CV TIER_1 English(EN) · Yong Man Ro · 2026-04-27 04:42

Robust Grounding with MLLMs against Occlusion and Small Objects via Language-guided Semantic Cues

While Multimodal Large Language Models (MLLMs) have enhanced grounding capabilities in general scenes, their robustness in crowded scenes remains underexplored. Crowded scenes entail visual challenges (i.e., occlusion and small objects), which impair object semantics and degrade …

报道来源 [2]

Robust Grounding with MLLMs against Occlusion and Small Objects via Language-guided Semantic Cues

Robust Grounding with MLLMs against Occlusion and Small Objects via Language-guided Semantic Cues

相关实体

相关话题