Speaker Diarization (SD) aims at grouping speech segments that belong to the
same speaker. This task is required in many speech-processing applications,
such as rich meeting transcription. In this context, distant microphone arrays
usually capture the audio signal. Beamforming, i.e., spatial filtering, is a
common practice to process multi-microphone audio data. However, it often
requires an explicit localization of the active source to steer the filter.
This paper proposes a self-attention-based algorithm to select the output of a
bank of fixed spatial filters. This method serves as a feature extractor for
joint Voice Activity (VAD) and Overlapped Speech Detection (OSD). The speaker
diarization is then inferred from the detected segments. The approach shows
convincing distant VAD, OSD, and SD performance, e.g. 14.5% DER on the
AISHELL-4 dataset. The analysis of the self-attention weights demonstrates
their explainability, as they correlate with the speaker's angular locations.

基于自注意力的算法用于选择固定空间滤波器的输出，从而用于联合语音活动检测和重叠说话检测的特征提取器，进而推断说话人分离。自注意力权重的分析表明其可解释性，因为它们与说话者的角度位置相关。

ASoBO：会议中远程说话者分辨的注意力波束选择

ASoBO: Attentive Beamformer Selection for Distant Speaker Diarization in  Meetings

We propose ViDeNN: a CNN for Video Denoising without prior knowledge on the
noise distribution (blind denoising). The CNN architecture uses a combination
of spatial and temporal filtering, learning to spatially denoise the frames
first and at the same time how to combine their temporal information, handling
objects motion, brightness changes, low-light conditions and temporal
inconsistencies. We demonstrate the importance of the data used for CNNs
training, creating for this purpose a specific dataset for low-light
conditions. We test ViDeNN on common benchmarks and on self-collected data,
achieving good results comparable with the state-of-the-art.

该研究提出了 ViDeNN: 一种用于视频去噪的卷积神经网络，能够在没有先验知识的情况下去除噪声，并使用组合的空间和时间滤波进行学习，处理物体运动、亮度变化、低光条件和时间不一致性。通过特定数据集的训练，ViDeNN 在常见基准测试和自采数据上取得了与现有技术相当的好结果。

ViDeNN: 深度盲视频降噪

ViDeNN: Deep Blind Video Denoising

This paper considers the problem of interference control through the use of
second-order statistics in massive MIMO multi-cell networks. We consider both
the cases of co-located massive arrays and large-scale distributed antenna
settings. We are interested in characterizing the low-rankness of users'
channel covariance matrices, as such a property can be exploited towards
improved channel estimation (so-called pilot decontamination) as well as
interference rejection via spatial filtering. In previous work, it was shown
that massive MIMO channel covariance matrices exhibit a useful finite rank
property that can be modeled via the angular spread of multipath at a MIMO
uniform linear array. This paper extends this result to more general settings
including certain non-uniform arrays, and more surprisingly, to two dimensional
distributed large scale arrays. In particular our model exhibits the dependence
of the signal subspace's richness on the scattering radius around the user
terminal, through a closed form expression. The applications of the
low-rankness covariance property to channel estimation's denoising and
low-complexity interference filtering are highlighted.

本文研究在巨量 MIMO 多小区网络中通过使用二阶统计量来控制干扰的问题，研究了共位巨型阵列和大规模分布式天线设置的情况，并着重研究了用户信道协方差矩阵的低秩特性，以期通过改进通道估计（所谓的导频净化）和空间滤波来抑制干扰。