发表机构
University of California, Riverside; California Institute of Technology; University of California; NASA Postdoctoral; The Graduate Center, CUNY; University of Illinois at Chicago（加州大学河滨分校; 加州理工学院; 加州大学; NASA博士后项目; 纽约市立大学研究生中心; 伊利诺伊大学芝加哥分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出深度学习框架，通过卷积神经网络提升WISE红外成像到斯皮策分辨率，在COSMOS领域验证，能降低通量误差、提高源峰值恢复率，减少过度平滑问题，为全天空超分辨率应用提供途径。
AI 中文摘要
我们提出了一个深度学习框架，该框架可实现从WISE W1（3.4微米）到斯皮策IRAC Ch1（3.6微米）的4.6倍空间超分辨率，并在COSMOS领域对其性能进行了表征。我们的样本包括在WISE/斯皮策重叠区域内均匀抽取的约390,000对裁剪图像，以及用于报告所有指标的83,592个裁剪图像的验证测试集。该框架使用卷积神经网络（增强残差通道注意力网络），通过强调在拥挤区域准确恢复源的损失函数进行训练。该模型在固定孔径中恢复中心源的总通量，中位数相对误差为11%，比插值基线好约2倍；在最暗的四分位数上增益达到约3倍。亮度依赖性是单调的：孔径积分通量误差从最暗四分位数的13%降至最亮四分位数的8%。在WISE混合而斯皮策分离源的3-5角秒间距处，该模型可恢复斯皮策真值中可检测到的35%的源峰值，而插值为9%。特征性失败模式是源轮廓过度平滑，这会使积分通量测量值向上偏置；这种模式在定性上与插值基线相似，但对于训练模型在定量上较小。这些结果表明真正的分辨率增强和源去混合，为在斯皮策无法覆盖的全天空区域应用超分辨率提供了一条途径。附录在W2 -> IRAC Ch2上重复了分析，结果一致；训练好的模型和代码可公开获取。
英文摘要
We present a deep-learning framework that performs 4.6x spatial super-resolution from WISE W1 (3.4 micron) toward Spitzer IRAC Ch1 (3.6 micron), and characterize its behavior on the COSMOS field. We train on ~390,000 paired cutouts drawn uniformly within the WISE/Spitzer overlap and report all metrics on a held-out test set of 83,592 cutouts. The framework uses a convolutional neural network trained with a loss function that emphasizes accurate recovery of sources in crowded fields. The model recovers the total flux of the central source in a fixed aperture to a median relative error of 11%, a factor of ~2 better than the interpolation baselines; the gain reaches ~3x on the faintest quartile. The brightness dependence is monotonic: the aperture integrated flux error decreases from 13% on the faintest quartile to 8% on the brightest. At the 3-5 arcsec separations where WISE blends sources that Spitzer separates, the model recovers 35% of the source peaks detectable in the Spitzer truth compared with 9% for interpolation. The characteristic failure mode is oversmoothing of source profiles, which biases integrated flux measurements upward; this pattern is qualitatively similar to that of the interpolation baselines but is quantitatively smaller for the trained model. These results suggest genuine resolution enhancement and source deblending, providing a path toward applying super-resolution across the all-sky area that Spitzer could not cover. An appendix replicates the analysis at W2 -> IRAC Ch2 with consistent results; the trained model and code are publicly available.
Comments28 pages, 14 figures, 10 tables. Submitted to AAS Journals
