Searcharxiv⌕ Search

arXiv subjects

Chengyan Zhang

Publications and source records attributed to Chengyan Zhang.

3 recordsLinked to original sources

Machine Learning the $GW$ Self-Energy Operator for Real Materials

Machine learning of $GW$ in solids has largely focused on prediction of scalar quasiparticle energy corrections. Here, we introduce a phase-aware many-body transformer that predicts the full $\mathbf{k}$-resolved complex self-energy matrix. As a proof of principle, we apply the model to predict quasiparticle energies before and after diagonalization of the quasiparticle Hamiltonian, obtaining a test set mean absolute error (MAE) of 29 meV for a dataset of 1745 2D materials in the static COHSEX approximation and 60 meV for a more limited dataset of 433 materials calculated within a plasmon-pole model. COHSEX pretraining followed by finetuning reduces this MAE to 42 meV. The predicted matrices also accurately reproduce QP wavefunctions.

cond-mat.mtrl-sci↗

$U(1)$ Gauge-Equivariant Representation Learning of Bloch States

Representation learning, or featurization, of mean-field Bloch states provides an important route for incorporating electronic information into machine-learning models of first-principles condensed-phase systems. A challenge in representing Bloch states is the gauge redundancy of each state. In this work, we propose to address this issue by explicitly incorporating gauge equivariance in the machine-learning framework. We develop a $U(1)$-equivariant autoencoder that compresses the plane-wave-basis Bloch states obtained from density functional theory into a low-dimensional latent representation. We achieve an average overlap of 0.970 between original wavefunctions and the reconstructions from a 64-dimensional latent space using a dataset of two-dimensional insulators. We further show that $U(1)$ equivariance endows the latent space with physically meaningful structures, including smoothness and topological information. Finally, we present a proof-of-principle application in which the latent representations are used to predict the off-diagonal elements of the $GW$ self energy matrix, and demonstrate the importance of enforcing gauge equivariance in the prediction.

physics.comp-ph↗

VR-based generation of photorealistic synthetic data for training hand-object tracking models

Supervised learning models for precise tracking of hand-object interactions (HOI) in 3D require large amounts of annotated data for training. Moreover, it is not intuitive for non-experts to label 3D ground truth (e.g. 6DoF object pose) on 2D images. To address these issues, we present "blender-hoisynth", an interactive synthetic data generator based on the Blender software. Blender-hoisynth can scalably generate and automatically annotate visual HOI training data. Other competing approaches usually generate synthetic HOI data compeletely without human input. While this may be beneficial in some scenarios, HOI applications inherently necessitate direct control over the HOIs as an expression of human intent. With blender-hoisynth, it is possible for users to interact with objects via virtual hands using standard Virtual Reality hardware. The synthetically generated data are characterized by a high degree of photorealism and contain visually plausible and physically realistic videos of hands grasping objects and moving them around in 3D. To demonstrate the efficacy of our data generation, we replace large parts of the training data in the well-known DexYCB dataset with hoisynth data and train a state-of-the-art HOI reconstruction model with it. We show that there is no significant degradation in the model performance despite the data replacement.

cs.CV↗