Abstract:In this paper, we introduce a Unet model of deep learning algorithms for reconstructions of the 3D peculiar velocity field, which simplifies the reconstruction process with enhanced precision. We test the adaptability of the Unet model with simulation data under more realistic conditions, including the redshift space distortion (RSD) effect and halo mass threshold. Our results show that the Unet model outperforms the analytical method that runs under ideal conditions, with a 16% improvement in precision, 13% in residuals, 18% in correlation coefficient and 27% in average coherence. The deep learning algorithm exhibits exceptional capacities to capture velocity features in non-linear regions and substantially improve reconstruction precision in boundary regions. We then apply the Unet model trained under SDSS observational conditions to the SDSS DR7 data for observational 3D peculiar velocity reconstructions.

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to solve the problem of reconstructing the 3D peculiar velocity field from observational data. Specifically, the author introduced a Unet model based on deep learning to simplify and improve the accuracy of the reconstruction of the peculiar velocity field. The following are the main problems and challenges of this research: 1. **Reconstruction accuracy**: - Although existing analytical methods can provide certain reconstruction results when operating under ideal conditions, they have limitations when dealing with non - linear regions and boundary regions, resulting in insufficient accuracy. - The reconstruction of the peculiar velocity field requires extracting information from the density field, and existing methods perform poorly when dealing with issues such as redshift space distortion (RSD) and halo mass thresholds. 2. **Data - driven methods**: - Traditional theory - driven algorithms are complex and computationally expensive, especially when dealing with large - scale cosmic structures. Therefore, researchers hope to simplify this process and improve efficiency and accuracy through machine learning, especially deep - learning algorithms. 3. **Challenges in practical applications**: - In actual observations, peculiar velocity measurements are affected by the uncertainty of distance modulus estimation, especially in the high - redshift region. This limits peculiar velocity surveys to a very small redshift range. - In contrast, redshift surveys (such as DESI) can cover a larger redshift range. Therefore, it is possible to reconstruct the velocity field from these data, but it is necessary to solve the problem of how to accurately reconstruct the velocity field from the overdense field. 4. **Improving existing methods**: - This research hopes to overcome the limitations of existing methods by introducing deep - learning algorithms, especially the Unet model, so as to achieve higher reconstruction accuracy under more complex conditions (such as RSD effects and halo mass thresholds). 5. **Verification and application**: - To verify the effectiveness of the model, the researchers used simulated data for training and verification, and finally applied the trained model to actual observational data (such as SDSS DR7) to evaluate its performance under real conditions. ### Summary The main goal of this paper is to simplify and improve the accuracy of reconstructing the 3D peculiar velocity field from observational data by introducing a Unet model based on deep learning, especially performing well when dealing with non - linear regions and boundary regions. At the same time, the research also explored the application potential of this model in actual observational data, showing its advantages over traditional analytical methods.

Peculiar Velocity Reconstruction From Simulations and Observations Using Deep Learning Algorithms

AI-assisted Reconstruction of Cosmic Velocity Field from Redshift-Space Spatial Distribution of Halos

A Calibration-Informed Deep Learning Model for Three-Dimensional Particle Reconstruction of Volumetric Particle Image Velocimetry

Cosmic Velocity Field Reconstruction Using AI

Neural network reconstruction of density and velocity fields from the 2MASS Redshift Survey

(DarkAI) Mapping the large-scale density field of dark matter using artificial intelligence

3D-EddyNet: A Novel Approach for Identifying Three-Dimensional Morphological Features of Mesoscale Eddies in the Ocean

kSZ Pairwise Velocity Reconstruction with Machine Learning

Kinematic Sunyaev-Zel′dovich pairwise velocity reconstruction with machine learning

Velocity reconstruction in the era of DESI and Rubin/LSST. I. Exploring spectroscopic, photometric, and hybrid samples

Reconstructing the cosmological density and velocity fields from redshifted galaxy distributions using V-net

Velocity reconstruction in the era of DESI and Rubin (part I): Exploring spectroscopic, photometric & hybrid samples

Estimation of line-of-sight velocities of individual galaxies using neural networks I. Modelling redshift-space distortions at large scales

UHRNet: A Deep Learning-Based Method for Accurate 3D Reconstruction from a Single Fringe-Pattern

Velocity recostruction with graph neural networks

Velocity reconstruction in the era of DESI and Rubin/LSST. II. Realistic samples on the light cone

Joint velocity and density reconstruction of the Universe with nonlinear differentiable forward modeling

Visual Geometry Group-UNet: Deep Learning Ultrasonic Image Reconstruction for Curved Parts.

UNeR3D: Versatile and Scalable 3D RGB Point Cloud Generation from 2D Images in Unsupervised Reconstruction

Velocity reconstruction in the era of DESI and Rubin (part II): Realistic samples on the light cone