International Business Machines Corporation

National Institute of Metrology, China

Beijing Bytedance Network Technology Co., Ltd.

China Mobile Communications Corporation

Lu Huang

Jun Hong Zhao

Hui Huang

Ang Yi

Ya Juan Tong

Xue Mei Deng

Wei Wang

Jun Zhang

Yang Liang

Chen Bo Li

Yan Yang

Yin Qian

Chen Li

Xiang Cai

Song Bai

Zhiheng Liu

Xue Mei D Deng

Yuqi Fang

Miao Sun

Xia Liu

Dongxue Dai

Sitian Gao

Disclosed are a method and apparatus for training an acoustic network model, and an electronic device, a readable storage medium, a computer program product and a program, wherein same belong to the technical field of data processing. The method comprises: acquiring a training sample for training of an acoustic network model, wherein the training sample includes one or more speech files (S); performing peak detection on speech included in the speech file by using a BLSTM network, so as to divide letters included in the speech file on the basis of a detected peak, thereby forming segmented speech (S); performing semantic prediction on the segmented speech on the basis of a preset CTC segmentation network to obtain a text prediction result (S); and after the prediction result meets a preset performance requirement, stopping the training of the acoustic network model so as to perform text prediction on input speech by using the acoustic network model (S). The method can improve the training efficiency of an acoustic network model.

Method and apparatus for training acoustic network model, and electronic device

The present disclosure relates to a field programmable gate array (FPGA)-based multi-channel dynamic light scattering (DLS) autocorrelation system and method. The system includes a DLS generation apparatus, a photon correlator, and a host computer, where the photon correlator includes an FPGA and a universal serial bus (USB) communication module; the DLS generation apparatus is connected to the FPGA; the FPGA is configured to count and perform correlation calculation on photon pulses generated by the DLS generation apparatus; the USB communication module is connected to the host computer; the FPGA includes a dual counter module and a correlation calculation module; the dual counter module is connected to the DLS generation apparatus and the correlation calculation module; the correlation calculation module is connected to the USB communication module; the dual counter module includes a plurality of dual counters; and the correlation calculation module includes a plurality of correlators.

FPGA-based multi-channel dynamic light scattering autocorrelation system and method

A method for achieving terminal-pair definition of four-terminal-pair (4TP) impedance and an application thereof are provided, which belongs to the field of precision measurement and metrology. A current output terminal of a two-stage follower is connected to a high current terminal of impedance through a coaxial line, and a voltage output terminal of the two-stage follower is connected to a high voltage terminal of the impedance through the coaxial line, which makes current of the high voltage terminal be 0, and core wire currents and outer wire currents of the high current terminal equal and reverse. The terminal-pair definition of the 4TP impedance can be satisfied; and the follower is added to make a bridge ratio variable and isolate effects of bridge load changes, thereby accelerating a balancing speed of the 4TP impedance bridge, and achieving accurate and fast comparative measurement having high precision of the 4TP AC impedance.

Method for achieving terminal-pair definition of four-terminal-pair impedance and application thereof

A mechanism is provided for enhancing security access to a data processing system. Responsive to authenticating a first factor of the multi-factor authentication by matching a scanned fingerprint/thumbprint of a user to at least one previously scanned fingerprint/thumbprint in a set of previously scanned fingerprints/thumbprints, a determination is made as to whether a touch pressure level exerted by the user on a touch pressure sensor associated with the data processing system matches at least one previously stored touch pressure level in a set of previously stored touch pressure levels. Responsive to matching the touch pressure level exerted by the user to a previously stored touch pressure level in the set of previously stored touch pressure levels, a second factor of the multi-factor authentication is authenticated. Access is then granted for the user to the data processing system.

Enhanced security access

In an approach, data of a plurality of points is sampled in a target area, wherein the data of each point of the plurality of points comprises position information and a height value. A first area of a target area is determined, wherein the height value of each point of the plurality of points in the first area complies with a first range. A second area of the target area is determined, wherein the height value of each point of the plurality of points in the second area complies with a second range. A third area of the target area is determined, wherein the height value of each point of the plurality of points in the third area complies with a third range. A first pattern is generated, wherein the first pattern is a combination of the first area and the third area.

Recognition for overlapped patterns

A method for replaying operations on a widget in a graphical user interface (GUI) includes acquiring, from an operation record, operation information related to the widget and a first unique identification (UID) of the widget. A widget having a second UID is determined, wherein the second UID is generated based on a call stack for creating the widget in the GUI. The second UID is the same as the first UID. It may be determined whether the widget on the GUI is operable. An operation is executed on the widget according to the operation information related to the widget.

Replaying operations on widgets in a graphical user interface

The present disclosure provides a communication method, a communication system, a resource pool management system, a switch device and a control device in a Software-Defined Network (SDN). The communication method may include steps of: obtaining network attributes of one or more virtual machines which are configured by a resource pool management system based on operations on a network resource pool by a user; converting tenant networking topology information derived from the network attributes into flow tables, in the case that a first virtual machine and a second virtual machine attempt to communicate with each other for the first time; and distributing the flow tables to one or more switch devices corresponding to the one or more virtual machines, so as to directly discard packets to be transmitted during the communication for the first time and subsequent communication between the first virtual machine and the second virtual machine, in response to determining by one of the switch devices that the communication between the first virtual machine and the second virtual machine is disabled based on a corresponding flow table. In the communication method, it is not necessary for the controller to broadcast the ARP messages to the whole SDN, and it is reduced the number of the packet_in packets to be transmitted to the SDN controller, so that the network resources of the SDN may be fully utilized, and the availability of the network resources and the controller resources are improved.

Communication method, communication system, resource pool management system, switch device and control device

Growing community of inventors

Beijing, China

Lu Huang