Abstract: In one aspect, a method of speech signal processing using artificial intelligence, includes receiving, with at least one processor, a digital speech signal, converting, with the at least one processor, the digital speech signal to text, performing self-alignment pre-training of an encoder on entities and their synonyms, label attention training the pretrained encoder by aligning label-text joint representation to that of label synonym-text joint representations, and identifying, with the trained encoder, entities in a span in the text.

Abstract: A horizontal sampling soil drilling rig and a control method are used to solve the problem that the soil sampling device in the prior art has difficulty in horizontal sampling. The horizontal sampling soil drilling rig includes a rack, a connection assembly, a drive mechanism, a spiral drill rod, and a sampler. The rack is provided with a spline shaft and a screw rod in parallel. The screw rod and the spline shaft are connected through a first transmission assembly in transmission connection. The drive mechanism includes a rotation driver, a second transmission assembly, and a spline shaft sleeve. The spline shaft sleeve is rotationally connected to the rack, and the spline shaft sleeve and the spline shaft mesh with each other. The rotation driver is connected with the spline shaft through the second transmission assembly in transmission connection.

Abstract: A processing unit can train a model as a joint multi-domain recurrent neural network (JRNN), such as a bi-directional recurrent neural network (bRNN) and/or a recurrent neural network with long-short term memory (RNN-LSTM) for spoken language understanding (SLU). The processing unit can use the trained model to, e.g., jointly model slot filling, intent determination, and domain classification. The joint multi-domain model described herein can estimate a complete semantic frame per query, and the joint multi-domain model enables multi-task deep learning leveraging the data from multiple domains. The joint multi-domain recurrent neural (JRNN) can leverage semantic intents (such as, finding or identifying, e.g., a domain specific goal) and slots (such as, dates, times, locations, subjects, etc.) across multiple domains.

Abstract: A first control unit sends a heartbeat message to a subscriber unit, the subscriber unit forwards the heartbeat message to a second control unit, and the second control unit makes a master/backup decision based on the heartbeat message.

Abstract: Systems and methods are disclosed for inquiry-based deep learning. In one implementation, a first content segment is selected from a body of content. The content segment includes a first content element. The first content segment is compared to a second content segment to identify a content element present in the first content segment that is not present in the second content segment. Based on an identification of the content element present in the first content segment that is not present in the second content segment, the content element is stored in a session memory. A first question is generated based on the first content segment. The session memory is processed to compute an answer to the first question. An action is initiated based on the answer. Using deep learning, content segments can be encoded into memory. Incremental questioning can serve to focus various deep learning operations on certain content segments.

Abstract: A processing unit can train a model as a joint multi-domain recurrent neural network (JRNN), such as a bi-directional recurrent neural network (bRNN) and/or a recurrent neural network with long-short term memory (RNN-LSTM) for spoken language understanding (SLU). The processing unit can use the trained model to, e.g., jointly model slot filling, intent determination, and domain classification. The joint multi-domain model described herein can estimate a complete semantic frame per query, and the joint multi-domain model enables multi-task deep learning leveraging the data from multiple domains. The joint multi-domain recurrent neural network (JRNN) can leverage semantic intents (such as, finding or identifying, e.g., a domain specific goal) and slots (such as, dates, times, locations, subjects, etc.) across multiple domains.

Abstract: Systems and methods are disclosed for inquiry-based deep learning. In one implementation, a first content segment is selected from a body of content. The content segment includes a first content element. The first content segment is compared to a second content segment to identify a content element present in the first content segment that is not present in the second content segment. Based on an identification of the content element present in the first content segment that is not present in the second content segment, the content element is stored in a session memory. A first question is generated based on the first content segment. The session memory is processed to compute an answer to the first question. An action is initiated based on the answer. Using deep learning, content segments can be encoded into memory. Incremental questioning can serve to focus various deep learning operations on certain content segments.

Abstract: Provided are a tumor immune enhancer, a pharmaceutical composition thereof, and a preparation method therefor. The enhancer comprises one or more polypeptides using sequences shown in SEQ ID NOs.: 1-9 as core structures, and can be used for preparing tumor vaccines.

Abstract: A system in package includes a memory-die stack in memory module that is stacked vertically with respect to a processor die. Each memory die in the memory-die stack includes a vertical bond wire that emerges from a matrix for connection. Some configurations include the vertical bond wire emerging orthogonally beginning from a bond-wire pad. The matrix encloses the memory-die stack, the spacer, and at least a portion of the processor die.

Abstract: A flexible substrate, a touch display substrate, manufacturing methods thereof, and a touch display device are provided. The flexible substrate includes: a first flexible base substrate; a first buffer layer arranged on the first flexible base substrate; a first pressure sensing electrode arranged on the first buffer layer; and a second flexible base substrate covering the first pressure sensing electrode. The first pressure sensing electrode is capable of forming a pressure sensing capacitor with a second pressure sensing electrode arranged at a side of the second flexible base substrate away from the first pressure sensing electrode.

Abstract: According to an aspect of an embodiment, a method may include obtaining a data set that includes categories (or features), and a target criteria. The method may further include obtaining a first decision tree model using the data set. The method may further include ranking the categories based on the first decision tree model and removing low-ranking categories from the data set. The method may further include generating a second decision tree model using the data set. The second decision tree model may include branch nodes. Each of branch nodes may represent a branch criteria. The method may further include pruning a branch node. The method may further include designating a remaining branch nodes as a rule node. The method may further include generating a rule based on the branch criteria of the rule node and presenting the rule in a graphical user interface.

Abstract: The present invention relates to a method for preparing a bactericidal phage vector, (pharmaceutical) compositions comprising such phage vectors, also for use in treating diseases, particularly those caused by (antimicrobial resistance) bacterial cells.

Abstract: A text adjustment method includes: obtaining a to-be-processed text; determining whether to adjust the to-be-processed text according to the to-be-processed text and a context text of the to-be-processed text; in response to determining to adjust the to-be-processed text, determining adjustment character information and adjustment position information of the to-be-processed text according to the context text; and determining an adjusted to-be-processed text according to the adjustment character information and the adjustment position information.

Abstract: The present invention relates to a method for simultaneously separating cannabidivarin and cannabigerol, comprising: sufficiently oscillating a solvent system, allowing the solvent system to stand, and separately collecting an upper phase and a lower phase; dissolving a commercially available industrial hemp full-spectrum refined oil into the upper phase, performing separation using high-speed countercurrent chromatography with the upper phase as a stationary phase and the lower phase as a mobile phase to obtain a mixture of cannabidivarin and the mobile phase and a mixture of cannabigerol and the mobile phase, respectively, and removing the mobile phase to obtain cannabidivarin and cannabigerol. The present invention uses high-speed countercurrent chromatography to simultaneously separate and purify to obtain cannabidivarin (CBDV) with the purity of greater than 98% and cannabigerol (CBG) with the purity of greater than 97% from the industrial hemp full-spectrum refined oil for the first time.

Abstract: A processing unit can extract salient semantics to model knowledge carryover, from one turn to the next, in multi-turn conversations. Architecture described herein can use the end-to-end memory networks to encode inputs, e.g., utterances, with intents and slots, which can be stored as embeddings in memory, and in decoding the architecture can exploit latent contextual information from memory, e.g., demographic context, visual context, semantic context, etc. e.g., via an attention model, to leverage previously stored semantics for semantic parsing, e.g., for joint intent prediction and slot tagging. In examples, architecture is configured to build an end-to-end memory network model for contextual, e.g., multi-turn, language understanding, to apply the end-to-end memory network model to multiple turns of conversational input; and to fill slots for output of contextual, e.g., multi-turn, language understanding of the conversational input.

Abstract: A processing unit can operate a first recurrent computational model (RCM) to provide first state information and a predicted result value. The processing unit can operating a first network computational model (NCM) to provide respective expectation values of a plurality of actions based at least in part on the first state information. The processing unit can provide an indication of at least one of the plurality of actions, and receive a reference result value, e.g., via a communications interface. The processing unit can train the first RCM based at least in part on the predicted result value and the reference result value to provide a second RCM, and can train the first NCM based at least in part on the first state information and the at least one of the plurality of actions to provide a second NCM.

Abstract: In some examples, a computing device refines feature information of query text. The device repeatedly determines attention information based at least in part on feature information of the image and the feature information of the query text, and modifies the feature information of the query text based at least in part on the attention information. The device selects at least one of a predetermined plurality of outputs based at least in part on the refined feature information of the query text. In some examples, the device operates a convolutional computational model to determine feature information of the image. The device network computational models (NCMs) to determine feature information of the query and to determine attention information based at least in part on the feature information of the image and the feature information of the query. Examples include a microphone to detect audio corresponding to the query text.

Abstract: Provided herein are cinchonium betaine catalysts and methods of promoting asymmetric butenolide isomerization reactions using the same.

Abstract: A method for bonding a three-dimensional cover, which includes the following steps: bonding a flexible display panel on a heavy release film through optical adhesive to obtain a bonded body; placing the bonded body on a carrying surface of a conformable tool, making the flexible display panel face the carrying surface of the conformable tool, and fixing two ends of the heavy release film in the bonded body to fixtures of the conformable tool; controlling the fixtures of the conformable tool to adjust the shape of the bonded body, so that the adjusted shape of the bonded body matches the three-dimensional cover; tearing off the heavy release film from the bonded body to obtain the flexible display panel with optical adhesive, and bonding the three-dimensional cover to the flexible display panel using the optical adhesive.

Abstract: In some examples, a computing device refines feature information of query text. The device repeatedly determines attention information based at least in part on feature information of the image and the feature information of the query text, and modifies the feature information of the query text based at least in part on the attention information. The device selects at least one of a predetermined plurality of outputs based at least in part on the refined feature information of the query text. In some examples, the device operates a convolutional computational model to determine feature information of the image. The device network computational models (NCMs) to determine feature information of the query and to determine attention information based at least in part on the feature information of the image and the feature information of the query. Examples include a microphone to detect audio corresponding to the query text.