From Images to Features: Unbiased Morphology Classification via Variational Auto-Encoders and Domain Adaptation
Authors:
Quanfeng Xu,
Shiyin Shen,
Rafael S. de Souza,
Mi Chen,
Renhao Ye,
Yumei She,
Zhu Chen,
Emille E. O. Ishida,
Alberto Krone-Martins,
Rupesh Durgesh
Abstract:
We present a novel approach for the dimensionality reduction of galaxy images by leveraging a combination of variational auto-encoders (VAE) and domain adaptation (DA). We demonstrate the effectiveness of this approach using a sample of low redshift galaxies with detailed morphological type labels from the Galaxy-Zoo DECaLS project. We show that 40-dimensional latent variables can effectively repr…
▽ More
We present a novel approach for the dimensionality reduction of galaxy images by leveraging a combination of variational auto-encoders (VAE) and domain adaptation (DA). We demonstrate the effectiveness of this approach using a sample of low redshift galaxies with detailed morphological type labels from the Galaxy-Zoo DECaLS project. We show that 40-dimensional latent variables can effectively reproduce most morphological features in galaxy images. To further validate the effectiveness of our approach, we utilised a classical random forest (RF) classifier on the 40-dimensional latent variables to make detailed morphology feature classifications. This approach performs similarly to a direct neural network application on galaxy images. We further enhance our model by tuning the VAE network via DA using galaxies in the overlapping footprint of DECaLS and BASS+MzLS, enabling the unbiased application of our model to galaxy images in both surveys. We observed that DA led to even better morphological feature extraction and classification performance. Overall, this combination of VAE and DA can be applied to achieve image dimensionality reduction, defect image identification, and morphology classification in large optical surveys.
△ Less
Submitted 13 October, 2023; v1 submitted 15 March, 2023;
originally announced March 2023.
Temporal variability of the telluric sodium layer
Authors:
D. S. Davis,
P. Hickson,
G. Herriot,
C. Y. She
Abstract:
The temporal variability of the telluric sodium layer is investigated by analyzing 28 nights of data obtained with the Colorado State University LIDAR experiment. The mean height power spectrum of the sodium layer was found to be well fit by a power law over the observed range of frequencies, 10 microhertz to 4 millhertz. The best fitting power law was found to be 10^βν^α, with α= -1.79 +/- 0.02…
▽ More
The temporal variability of the telluric sodium layer is investigated by analyzing 28 nights of data obtained with the Colorado State University LIDAR experiment. The mean height power spectrum of the sodium layer was found to be well fit by a power law over the observed range of frequencies, 10 microhertz to 4 millhertz. The best fitting power law was found to be 10^βν^α, with α= -1.79 +/- 0.02 and β= 1.12 +/- 0.40. Applications to wavefront sensing require knowledge of the behavior of the sodium layer at kHz frequencies. Direct measurements at these frequencies do not exist. Extrapolation from low-frequency behavior to high frequencies suggests that this variability may be a significant source of error for laser-guide-star adaptive optics on large-aperture telescopes.
△ Less
Submitted 11 September, 2006;
originally announced September 2006.
