研究人员开发了REALM，一个新颖的跨模态框架，旨在将RGB和事件相机数据对齐到一个共享的潜在流形中。该方法将事件表示投影到预训练的RGB基础模型的潜在空间中，并利用低秩自适应（LoRA）来弥合模态差距。REALM能够将图像训练的解码器零样本应用于事件流，用于深度估计和语义分割等任务，并在宽基线特征匹配中取得了最先进的成果。 AI
影响 能够实现图像训练模型到事件相机的零样本迁移，可能扩展其在机器人和自主系统中的应用。
排序理由 介绍新跨模态感知框架的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2605.00271v1 Announce Type: new Abstract: Event cameras provide several unique advantages over standard frame-based sensors, including high temporal resolution, low latency, and robustness to extreme lighting. However, existing learning-based approaches for event processing…
Event cameras provide several unique advantages over standard frame-based sensors, including high temporal resolution, low latency, and robustness to extreme lighting. However, existing learning-based approaches for event processing are typically confined to narrow, task-specific…