Abstract: A speech recognition method is provided. The method may include: obtaining speech data and a speech recognition result of the speech data, the speech data including speech of a plurality of speakers, and the speech recognition result including a plurality of words; determining speaking time of each of the plurality of speakers by processing the speech data; determining, based on the speaking times of the plurality of speakers and the speech recognition result, a corresponding relationship between the plurality of words and the plurality of speakers; determining, based on the corresponding relationship, at least one conversion word from the plurality of words, each of the at least one conversion word corresponding to at least two of the plurality of speakers; and re-determining the corresponding relationship between the plurality of words and the plurality of speakers based on the at least one conversion word.

Abstract: The present invention discloses an intelligent ward round method and system based on an SSVEP electroencephalogram signal, the method including: the interactive terminal is used to present a visual stimulus source that includes a plurality of candidate intention options to the patient, the acquisition terminal is used to collect the patient's electroencephalogram signal, the electroencephalogram signal obtained by the acquisition terminal is transmitted to the interactive terminal for electroencephalogram signal recognition, the patient's expression intention is determined, and the expression intentions comprise disease intentions, privacy intentions, psychological intentions, physiological intentions, environmental intentions, and safety intentions, and the intelligent ward round is completed. This method greatly improving the ability of communication between the patient and medical staff.

Abstract: The present invention discloses a dyskinesia rehabilitation training method and system based on electroencephalogram signal recognition; the method includes: the convolutional neural network is used to recognize the electroencephalogram signal, analyze the patient's motor intention, convert the motor intention into a control instruction for the exoskeleton device, drive the patient's limb movement by controlling the movement of the exoskeleton device, and assist the patient to complete the movement disorder rehabilitation training. In this method, the self-attention mechanism and self-distillation training were used to establish an electroencephalogram signal recognition model, analyzing the patient's motor intention to help motor nerve remodeling, which broke the passive and single problem of traditional rehabilitation methods, realized active rehabilitation of patients, and significantly improved the rehabilitation efficacy.

Abstract: An in-situ stress measurement device for ultra-deep non-vertical drilling adopts a technical route of single channel and bidirectional rolling of in-hole equipment, and its related functions are mainly realized by a bidirectional rolling device and a split-type axial-force-free circuit switching valve. The bidirectional rolling device can drive the in-hole equipment to move freely in a circumferential direction and back and forth, which changes sliding friction between the in-hole equipment and a hole wall into rolling friction during an in-situ stress measurement of non-vertical drilling. The split-type axial-force-free circuit switching valve provides a technical solution for a single-pipe test in the non-vertical drilling.

Abstract: The present application provides a cosmetic mirror comprising a first mirror body part and a second mirror body part in an opening/closing connection, wherein front surfaces of the first mirror body part and the second mirror body part are respectively provided with a first mirror surface and a second mirror surface, and a back surface of the first mirror body part is provided with an exposed third mirror surface.

Abstract: A comic image generating method, a computer device and a storage medium, wherein the method includes: acquiring target objects of a plurality of categories in a comic storyboard, wherein the plurality of categories at least include a character category, a scene category and a prop category; determining, for the target object of each category, a target image matching the target object according to an object level of the target object, wherein the object level is determined according to a degree of importance of the target object, and target images corresponding to the target objects with different object levels are determined through different image determination modes; and combining the target images matching the target objects according to image layer levels corresponding to the plurality of categories, to obtain a comic image corresponding to the comic storyboard.

Abstract: The present disclosure provides a comic image generation method and apparatus, a computer device, and a storage medium, and the method includes: acquiring pose text information corresponding to a comic storyboard, the pose text information being used for describing target action poses of at least one target object; determining a pose assistance image corresponding to the comic storyboard according to the pose text information, reference action poses of a reference object in the pose assistance image matching the target action poses; and generating comic images corresponding to the comic storyboard using an artificial intelligence model according to the pose text information and the pose assistance image, target objects in the comic images having the target action poses.

Abstract: The present disclosure discloses a method for synthesizing a speech. The method includes generating the speech based on a text with a speech synthesis model, wherein the speech synthesis model includes an embedding layer, a speech synthesis layer, and a position layer; and training the speech synthesis model when an evaluation index meets a preset condition, wherein the evaluation index includes one or more quality indexes determined based on at least a part of the text and at least a part of the speech.

Abstract: A method of generating a comic image, a computer device and a storage medium are provided. The method includes: acquiring a target novel to be used to generate a comic; determining keyword information corresponding to comic storyboards that correspond to the target novel in a plurality of comic image generation dimensions according to a content of the target novel; with respect to any comic storyboard of the comic storyboards, determining target model input information corresponding to the comic storyboard according to a mapping relationship library between dimension keywords and model input information, and the keyword information of the comic storyboards in the comic image generation dimensions, wherein the dimension keywords include a keyword that has been determined in any comic image generation dimension; and using an artificial intelligence model to generate a comic image corresponding to the comic storyboard according to the target model input information.

Abstract: A system and method for detecting trending topics in customer inquiries includes retrieving customer inquiries from a plurality of data sources for a target time window and reference time windows and detecting trending keywords in the target time window as compared to keywords in the reference time windows. Responsive to detecting the trending keywords, customer inquiries in the target time window that include one or more of the trending keywords are collected and a weight is measured for each collected customer inquiry based on weights of detected trending keywords in each collected customer inquiry. A connection graph is generated for the detected trending keywords and the collected customer inquiries. The detected trending keywords are then clustered into a plurality of trending topics based on the connection graph, and the trending topics are ranked based on the measured weights of the collected customer inquiries associated with each trending topic.

Abstract: A communication chip includes a switching (SW) die and a plurality of network processing (NP) dies. Any one of the plurality of NP dies is configured to: receive a first packet outside the communication chip through an external port, obtain destination information of the first packet, and send a second packet including the first packet and the destination information through an internal port. The destination information indicates a destination NP die of the first packet. The SW die is configured to: receive the second packet, and send the second packet to a first NP die in the plurality of NP dies based on the destination information. The first NP die is configured to: receive the second packet, and send the second packet outward through the external port.

Abstract: A speech recognition method is provided. The method may include: obtaining speech data and a speech recognition result of the speech data, the speech data including speech of a plurality of speakers, and the speech recognition result including a plurality of words; determining speaking time of each of the plurality of speakers by processing the speech data; determining, based on the speaking times of the plurality of speakers and the speech recognition result, a corresponding relationship between the plurality of words and the plurality of speakers; determining, based on the corresponding relationship, at least one conversion word from the plurality of words, each of the at least one conversion word corresponding to at least two of the plurality of speakers; and re-determining the corresponding relationship between the plurality of words and the plurality of speakers based on the at least one conversion word.

Abstract: The present disclosure relates to the technical field of space electric propulsion of spacecrafts, and discloses a thrust-quantitatively-controllable and self-neutralizable Kaufman ion thruster and a use method thereof. The Kaufman ion thruster includes a discharge chamber, a central cathode, a gas supply assembly, a steel magnetic assembly, an insulating barrier and a grid system, wherein the central cathode is coaxially inserted in the center of a front panel; the gas supply assembly includes an electronegative working substance gas source, a conventional working substance gas source, N electronegative working medium gas supply pipes and N working medium gas supply pipes; the insulating barrier includes a central connecting rod and 2N fins. Self-neutralization can be achieved without the need of neutralizers. In addition, the center-oriented thrust modes can be used for spacecraft orbit control; and the eccentric thrust mode can be used for spacecraft attitude adjustment.

Abstract: By one approach, a control circuit accesses planning computed tomography image information for a particular patient. The control circuit also accesses two-dimensional digital subtraction angiography image information for the particular patient. The control circuit then fuses the planning computed tomography image information with the two-dimensional digital subtraction angiography image information to provide fused image information of the patient target volume.

Abstract: The invention features a method for treating graft versus host disease in a human, including administering to said human a therapeutically effective amount of a cell including a chimeric antigen receptor (CAR) which is specifically directed against an immune checkpoint molecule.

Abstract: An in-situ stress measurement device for ultra-deep non-vertical drilling adopts a technical route of single channel and bidirectional rolling of in-hole equipment, and its related functions are mainly realized by a bidirectional rolling device and a split-type axial-force-free circuit switching valve. The bidirectional rolling device can drive the in-hole equipment to move freely in a circumferential direction and back and forth, which changes sliding friction between the in-hole equipment and a hole wall into rolling friction during an in-situ stress measurement of non-vertical drilling. The split-type axial-force-free circuit switching valve provides a technical solution for a single-pipe test in the non-vertical drilling.

Abstract: A terminal device, a camera, and a photographing apparatus are provided. The terminal device includes a first board and a first camera; the first board includes a light-transmitting area; the first camera includes a prism, a lens group, and a photosensitive element, the prism includes an object side and an image side, light transmitting through the light-transmitting area sequentially pass through the object side of the prism, the image side of the prism, and the lens group to irradiate onto the photosensitive element, a field of view of the light that the photosensitive element can receive is within a preset value range, and an optical axis of the lens group is not parallel to a vertical line of the light-transmitting area; and the preset value range is greater than or equal to 55°.

Abstract: Computerized systems and methods for segment based approach to routing picking are disclosed. The systems and methods may include performing steps for: receiving a floorplan of a first set of location IDs, wherein the first set of location IDs correspond to locations of multiple inventory items arranged in a floor; generating one or more base segments that connect the first set of location IDs; generating one or more route segments by combining the one or more base segments with one or more demand points corresponding to a second set of location IDs; generating one or more dispatch routes through the one or more route segments based on an optimal routing of resources that maximizes a density metric of the multiple inventory items included in the one or more dispatch routes; and assigning a first user to a combination of the one or more dispatch routes.

Abstract: The present disclosure discloses a method for synthesizing a speech. The method includes generating the speech based on a text with a speech synthesis model, wherein the speech synthesis model includes an embedding layer, a speech synthesis layer, and a position layer; and training the speech synthesis model when an evaluation index meets a preset condition, wherein the evaluation index includes one or more quality indexes determined based on at least a part of the text and at least a part of the speech.

Abstract: The present disclosure provides a processing method and apparatus for recovering a beam. The method includes: generating a first type of signaling comprising an index corresponding to a resource selected from a set of resources for a reference signal; transmitting the first type of signaling over a random access channel to a base station; assuming a Quasi Co-Location (QCL) relation between the selected resource and a downlink control channel, and the selected resource and a downlink data channel; receiving a second type of signaling carried in the downlink control channel using the assumed QCL relation between the selected resource and the downlink control channel, the second type of signaling further comprising an identification of the terminal device; and receiving a data transmission carried in the downlink data channel according to the QCL relation between the selected resource and the downlink data channel.