Cross attention layer, attention + FFN like Transformer's decoder.

Applies the anti-rectifier operation to the input tensor. Args: ----- X (torch.Tensor): Input tensor. Returns:

将整个二维数组归一化到0-1

最小-最大归一化到0-1范围

为每个worker初始化独立的RNG

Calculate covariance matrix of data

Residual Block

Model prints with number of trainable parameters

Return an activation function given a string

Computes the autocorrelation matrix for a given lag of the input samples. Args: ----- X (torch.Tensor): Samples matrix input wit

Finds the roots of a polynomial defined by its coefficients. equivalent to src.utils.find_roots, but support Pytorch. Args: coefficie

Multiply a matrix Kx with its Hermitian conjecture (gram matrix), and adds eps to the diagonal values of the matrix, ensuring a Hermit

Sets the differentiable subspace method for training subspaceNet. Options: "root_music", "esprit" Args: -----

Calculates the sum of diagonals in a square matrix. equivalent sum_of_diag, but support Pytorch. Args: matrix (torch.Tensor): Square

This is a more standard version of the position embedding, very similar to the one used by the Attention is all you need paper, genera

Residual Block Args: in_channels: number of input channels out_channels: number of output channels stride

ResNet like architecture for spatial feature extraction.

Deep Learning based Spatial Spectrum Estimator.

Reference: Zheng, Shilian, Zhuang Yang, Weiguo Shen, Luxin Zhang, Jiawei Zhu, Zhijin Zhao, and Xiaoniu Yang. “Deep Le

Reference: Merkofer, Julian P., Guy Revach, Nir Shlezinger, Tirza Routtenberg, and Ruud J. G. van Sloun. “DA-MUSIC: D

Initializes the SubspaceNet model. Args: ----- tau (int): Number of auto-correlation lags. M (int): Number of

CNN based DOA estimation network Args: num_out_grids (int): number of angle girds used References: Papageorg

Model prints with number of trainable parameters

为每个worker初始化独立的RNG

Implementation of the model-based Esprit algorithm, support Pytorch, intended for MB-DL models. the model sets for nominal and ideal condition