Disk Wind Feedback from High-mass Protostars. V. Application of Multi-Modal Machine Learning to Characterize Outflow Properties

TL;DR

This work tackles projection biases in protostellar outflow studies by introducing a multi-modal cross-attention framework that fuses spatial morphology and spectral kinematics. Trained on synthetic ALMA-like data derived from 3D MHD simulations, the model predicts outflow mass, inclination, and position angle with a Gaussian uncertainty model, leveraging Vision Transformers for robust performance. Interpretability analyses using Grad-CAM++, Integrated Gradients, and Occlusion Sensitivity reveal that central outflow regions and cavity walls drive predictions, aligning with physical expectations that morphology is the dominant diagnostic. When applied to real ALMA data, the framework yields stable position angle estimates and calibrated uncertainties, demonstrating a scalable, interpretable approach to overcoming projection effects in high-mass star formation studies and enabling automated analysis of large surveys.

Abstract

Characterizing protostellar outflows is fundamental to understanding star formation feedback, yet traditional methods are often hindered by projection effects and complex morphologies. We present a multi-modal deep learning framework that jointly leverages spatial and spectral information from CO observations to infer outflow mass, inclination, and position angle ($PA$). Our model, trained on synthetic ALMA observations generated from 3D magnetohydrodynamic simulations, utilizes a cross-attention fusion mechanism to integrate morphological and kinematic features with probabilistic uncertainty estimation. Our results demonstrate that Vision Transformer architectures significantly outperform convolutional networks, showing remarkable robustness to reduced spatial resolution. Interpretability analysis reveals a physically consistent hierarchy: spatial features dominate across all parameters, whereas spectral profiles provide secondary constraints for mass and inclination. Applied to observational ALMA data, the framework delivers stable mass and $PA$ estimates with exceptionally tightly constrained inclination angles. This study establishes multi-modal deep learning as a powerful, interpretable tool for overcoming projection biases in high-mass star formation studies.

Figures (22)

• Figure 1: Examples of synthetic outflows with different protostellar masses and inclination angles. Blue represents the blue-shifted outflow lobes ($-50<v_{blue}<-3$ km s$^{-1}$), while red represents the red-shifted lobes ($3<v_{red}<50$ km s$^{-1}$).
• Figure 2: Examples of synthetic outflow spectra for different protostellar masses and inclination angles. The upper panels show the spectra in linear scale, while the lower panels display them in logarithmic scale.
• Figure 3: Illustration of the multi-modal machine learning architecture. Image and spectrum encoders extract spatial and spectral features, which are fused via a cross-attention module to produce parameter predictions with associated uncertainties.
• Figure 4: Performance of the ViT_L_16 test model. See Appendix \ref{['Comparison of Performance for the Four Models']} (Figure \ref{['fig.resnet_vit_12msun_test']}) for a comparison of other model architectures. This model is trained on a restricted dataset that excludes samples with a protostellar mass of 12 $M_\odot$ and five specific inclination angles. Model performance is evaluated on outflows corresponding to the excluded mass and inclination angles. The top row compares the ground-truth values with the model predictions for protostellar mass, inclination angle, and position angle from left to right. The bottom row shows the probability distribution functions (PDFs) of the prediction errors, defined as the difference between the predicted and true values.
• Figure 5: Performance of the ViT_L_16 model, but trained on the full range of protostellar masses and inclination angles.