The classification of galaxy morphologies is an important step in the
investigation of theories of hierarchical structure formation. While human
expert visual classification remains quite effective and accurate, it cannot
keep up with the massive influx of data from emerging sky surveys. A variety of
approaches have been proposed to classify large numbers of galaxies; these
approaches include crowdsourced visual classification, and automated and
computational methods, such as machine learning methods based on designed
morphology statistics and deep learning. In this work, we develop two novel
galaxy morphology statistics, descent average and descent variance, which can
be efficiently extracted from telescope galaxy images. We further propose
simplified versions of the existing image statistics concentration, asymmetry,
and clumpiness, which have been widely used in the literature of galaxy
morphologies. We utilize the galaxy image data from the Sloan Digital Sky
Survey to demonstrate the effective performance of our proposed image
statistics at accurately detecting spiral and elliptical galaxies when used as
features of a random forest classifier.

发展了两种新颖的星系形态统计方法，并简化了现有的图像统计方法，通过随机森林分类器准确检测螺旋和椭圆星系。

螺旋 - 椭圆自动星系形态分类从望远镜图像

Spiral-Elliptical automated galaxy morphology classification from  telescope images

Recent advancements in areas such as natural language processing and computer
vision rely on intricate and massive models that have been trained using vast
amounts of unlabelled or partly labeled data and training or deploying these
state-of-the-art methods to resource constraint environments has been a
challenge. Galaxy morphologies are crucial to understanding the processes by
which galaxies form and evolve. Efficient methods to classify galaxy
morphologies are required to extract physical information from modern-day
astronomy surveys. In this paper, we introduce methods to learn from less
amounts of data. We propose using a hybrid transformer-convolutional
architecture drawing much inspiration from the success of CoAtNet and MaxViT.
Concretely, we use the transformer-convolutional hybrid with a new stack design
for the network, a different way of creating a relative self-attention layer,
and pair it with a careful selection of data augmentation and regularization
techniques. Our approach sets a new state-of-the-art on predicting galaxy
morphologies from images on the Galaxy10 DECals dataset, a science objective,
which consists of 17736 labeled images achieving $94.86\%$ top-$1$ accuracy,
beating the current state-of-the-art for this task by $4.62\%$. Furthermore,
this approach also sets a new state-of-the-art on CIFAR-100 and Tiny ImageNet.
We also find that models and training methods used for larger datasets would
often not work very well in the low-data regime. Our code and models will be
released at a later date before the conference.

本文提出了使用混合变压器卷积结构及选择数据增强和正则化技术的方法学习更少量的数据以从现代天文调查中提取物理信息，此方法在 Galaxy10 DECals 数据集上取得了 94.86% 的精度，在 CIFAR-100 和 Tiny ImageNet 上也实现了新的最佳效果。

Astroformer：对于分类任务仅仅增加数据并不一定足够

Astroformer: More Data Might Not be All You Need for Classification

We apply the spectral energy distribution (SED) fitting code ProSpect to
multiwavelength imaging for $\sim$7,000 galaxies from the GAMA survey at
$z<0.06$, in order to extract their star formation histories. We combine a
parametric description of the star formation history with a closed-box
evolution of metallicity where the present-day gas-phase metallicity of the
galaxy is a free parameter. We show with this approach that we are able to
recover the observationally determined cosmic star formation history (CSFH), an
indication that stars are being formed in the correct epoch of the Universe, on
average, for the manner in which we are conducting SED fitting. We also show
the contribution to the CSFH of galaxies of different present-day visual
morphologies, and stellar masses. Our analysis suggests that half of the mass
in present-day elliptical galaxies was in place 11 Gyr ago. In other
morphological types, the stellar mass formed later, up to 6 Gyr ago for
present-day irregular galaxies. Similarly, the most massive galaxies in our
sample were shown to have formed half their stellar mass by 11 Gyr ago, whereas
the least massive galaxies reached this stage as late as 4 Gyr ago (the
well-known effect of "galaxy downsizing"). Finally, our metallicity approach
allows us to follow the average evolution in gas-phase metallicity for
populations of galaxies, and extract the evolution of the cosmic metal mass
density in stars and in gas, producing results in broad agreement with
independent, higher-redshift observations of metal densities in the Universe.

该研究采用 ProSpect SED 拟合代码，结合一个针对金属参数的参数化恒星形成历史描述来恢复宇宙恒星形成历史，同时分析不同类型的星系对恒星形成历史和宇宙金属质量密度的贡献。