标题： 将多模态学习应用于早期分类瞬态检测（AppleCiDEr）I：数据集、方法和基础设施 显示英文标题

标题： Applying multimodal learning to Classify transient Detections Early (AppleCiDEr) I: Data set, methods, and infrastructure

摘要： 现代时域调查如Zwicky瞬变设施（ZTF）和空间与时间遗产调查（LSST）生成数十万到数百万个警报，需要对瞬变源和变星进行自动统一分类，以便有效后续跟进。 我们提出AppleCiDEr（将多模态学习应用于早期瞬变检测分类），一种新的框架，整合四种关键数据模态（测光、图像切片、元数据和光谱），以克服单一模态分类方法的局限性。 我们的架构引入了（i）两个用于测光的变压器编码器，（ii）一个具有领域专用元数据塔和专家混合融合的多模态卷积神经网络（CNN），用于结合元数据和图像，以及（iii）一个用于光谱分类的CNN。 在约30,000个真实ZTF警报上进行训练，AppleCiDEr实现了高精度，允许早期识别并建议对稀有瞬变光谱进行后续跟进。 该系统提供了首个使用真实观测数据对瞬变源和变星进行统一分类的框架，并可无缝集成到中介管道中，展示了其适用于LSST时代的准备情况。 显示英文摘要

摘要： Modern time-domain surveys like the Zwicky Transient Facility (ZTF) and the Legacy Survey of Space and Time (LSST) generate hundreds of thousands to millions of alerts, demanding automatic, unified classification of transients and variable stars for efficient follow-up. We present AppleCiDEr (Applying Multimodal Learning to Classify Transient Detections Early), a novel framework that integrates four key data modalities (photometry, image cutouts, metadata, and spectra) to overcome limitations of single-modality classification approaches. Our architecture introduces (i) two transformer encoders for photometry, (ii) a multimodal convolutional neural network (CNN) with domain-specialized metadata towers and Mixture-of-Experts fusion for combining metadata and images, and (iii) a CNN for spectra classification. Training on ~ 30,000 real ZTF alerts, AppleCiDEr achieves high accuracy, allowing early identification and suggesting follow-up for rare transient spectra. The system provides the first unified framework for both transient and variable star classification using real observational data, with seamless integration into brokering pipelines, demonstrating readiness for the LSST era.