Abstract: System, method, and computer readable medium for dynamic graph representation learning with self-supervision, including extracting a time window of data from the dynamic graph representation to obtain a history graph that represents a sub-set of the dynamic graph representation; generating, using an encoder model configured by a set of learned encoder parameters and implemented by the computer system, a set of embeddings for the history graph; and predicting, using a first decoder model configured by a first set of learned decoder parameters and implemented by the computer system, one or more predictions for the dynamic graph representation corresponding to the specific prediction task.

Abstract: An audio signal enhancement method includes: acquiring an audio signal; estimating a signal-to-noise ratio (SNR) of an audio frame of the audio signal; determining a SNR threshold for the audio frame; selecting an audio signal processing technique according to a comparison of the SNR threshold to the estimated SNR of the audio frame; filtering the audio frame using a Wiener filter applying the selected signal processing technique; and outputting the audio frame filtered using the Wiener filter applying the selected signal processing technique. A first-in, first-out (FIFO) signal processing technique is selected when the estimated SNR of the audio frame is less than the SNR threshold, and a log-energy voice activity detection (VAD) signal processing technique is selected when the estimated SNR of the audio frame is greater than the SNR threshold.

Abstract: An audio signal enhancement method includes: acquiring an audio signal; estimating a signal-to-noise ratio (SNR) of an audio frame of the audio signal; determining a SNR threshold for the audio frame; selecting an audio signal processing technique according to a comparison of the SNR threshold to the estimated SNR of the audio frame; filtering the audio frame using a Wiener filter applying the selected signal processing technique; and outputting the audio frame filtered using the Wiener filter applying the selected signal processing technique. A first-in, first-out (FIFO) signal processing technique is selected when the estimated SNR of the audio frame is less than the SNR threshold, and a log-energy voice activity detection (VAD) signal processing technique is selected when the estimated SNR of the audio frame is greater than the SNR threshold.

Abstract: A method includes: acquiring speech signals in a vehicle; dividing the speech signals into speech segments including one or more speech samples; processing a set of the speech segments using dual Kalman filters; and synthesizing the processed speech segments to construct noise-reduced speech signals. Each dual Kalman filter includes a first Kalman filter and a second Kalman filter, each speech segment in the set is processed using a different dual Kalman filter, and each speech segment in the set is processed in parallel with one another.

Abstract: A method, in a network node, for mapping a wholesale network user's identity onto a mobile network user identity that is routable in the mobile network, comprises: receiving, through a communication interface of the network node, a request from the user of the wholesale network for a service provided by the mobile network, where the user of the wholesale network is not registered in the mobile network. The method further comprises, based on the requested service, selecting a routable identity among a group of identities associated with the requested service, in the mobile network, by a processor of the network node. Finally, the method comprises: mapping the selected routable identity of the mobile network to the wholesale network user's identity, by the processor of the network node.