Abstract: A threat detection system for a mobile communication system, and a global device and a local device thereof are provided. The threat detection system is used for detecting and defensing low and slow distributed denial-of-service (LSDDoS) attacks. The global device is located in a core network of the mobile communication system, and is used for training a tensor neural network (TNN) model to build a threat classifier. The threat classifier is used for the local device to identify a plurality of threat types. The local device inputs the to-be-identified data into the threat classifier to generate a classification result corresponding to one of the threat types.

Abstract: A threat detection apparatus and a threat detection method thereof for a wireless communication system are provided. The threat detection apparatus receives an observed time difference of arrival (OTDOA) message for positioning a user equipment (UE) from a serving base station (BS), and determines that the UE connects to a false BS when the identity of the serving BS is not on the identity list. If the identity of the serving BS is on the identity list, the threat detection apparatus calculates a first distance according to the measurement report message transmitted from the UE, and calculates a second distance between the UE and the serving BS according to the OTDOA message. When the difference between the first distance and the second distance is larger than a threshold, the threat detection apparatus determines that the UE connects to the false BS.

Abstract: A networking behavior detector and a networking behavior detection method thereof for an indoor space are provided. The networking behavior detector receives a plurality of radio frequency (RF) signals in the indoor space and converts the RF signals to a plurality of digital signals. Next, the networking behavior detector calculates an energy value of each digital signal and filters out the digital signal, the energy value of which is smaller than a threshold, of the digital signals to generate an analysis signal. Finally, the networking behavior detector retrieves a plurality of energy feature values of each analysis signal to generate a feature datum, and analyzes the feature data through an identification model to generate an identification result. The identification result corresponds to one of a plurality of networking behaviors.

Abstract: A base station and a user equipment with enhanced PDCP duplication are provided. The base station generates at least one RRC message for a user equipment, wherein the at least one RRC message includes a set of multiple activation criteria, and each activation criterion corresponds to one of multiple RLC entities. The base station transmits the at least one RRC message to the user equipment. Multiple logical channels are provided in a MAC layer, each RLC entity corresponds to at least one of the logical channels, each RLC entity corresponds to a working status (activation or deactivation). At least one of the working statuses of the RLC entities is changed due to at least one of the activation criteria being satisfied, and the base station changes a transmission restriction of at least one specific logical channel among the logical channels based on the changed working statuses of the RLC entities.

Abstract: A missing data compensation method, missing data compensation system and non-transitory computer-readable medium are provided in this disclosure. The method includes the following operations: inputting a sensing signal by a sensor; searching for a historical data sections similar to a first data section from the plurality of historical data sections to generate a plurality of candidate data sections; calculating a plurality of data relation diagrams according to the first data section and the candidate data sections, respectively; utilizing a feature recognition model to calculate a plurality of similarity values according to the data relation diagrams; selecting a candidate data section corresponding to the maximum similarity value as a sample data section; and utilizing the data in the sample data section to compensate the data in the first data section to generate compensated data section.

Abstract: An image labeling apparatus, method, and computer program product are provided. The image labeling apparatus generates several image features of several to-be-processed images and classifies them into several groups. For each group, the image labeling apparatus: (a) selects the image feature closest to the center as a representative feature and a compared feature, (b) selects the image feature farthest to the compared feature as a candidate feature, and (c) compares the similarity between the candidate feature and each representative feature with a threshold. When all the similarities are smaller than the threshold, adopts the candidate feature as another representative feature and the compared feature and repeats (b) and (c). The group(s) whose number of representative features is more than another threshold will be re-classified.

Abstract: A cargo container image capturing method and system applied to a bridge crane are provided. The cargo container image capturing method includes the following steps. A human-machine interface (HMI) image of an HMI of the bridge crane is captured. Several values are obtained from the HMI image using an image recognition algorithm. At least one camera is controlled to capture a cargo container image according to the values.

Abstract: A data filtering system, a data selection method, and a state prediction system are provided. The state prediction system includes the data filtering system and a predictive model generation system. The data filtering system includes a data pre-processing device and a property selection device. The data pre-processing device transforms first sample data corresponding to a first detection property into first feature parameters, transforms second sample data corresponding to a second detection property into second feature parameters, and transforms third sample data corresponding to a third detection property into third feature parameters. The property selection device selects at least two of the detection properties according to the first feature parameters, the second feature parameters, and the third feature parameters. Then, the predictive model generation system trains a predictive model based on the at least two detection properties selected by the property selection device.

Abstract: A guidance system and method for action postures are provided. The guidance system includes a first skeleton recognition model, an editing device and a storage device. The first skeleton recognition model is used for generating a standard skeleton video according to a standard action posture video. The standard skeleton video includes a standard specific action posture of a person, and the standard skeleton video includes several skeleton pictures for standard specific action posture corresponding to the standard specific action posture. The editing device is used for receiving a command of an editor to generate at least one action posture guidance information corresponding to the skeleton pictures for standard specific action posture. The storage device is used for storing the skeleton pictures for standard specific action posture and the at least one action posture guidance information corresponding to the skeleton pictures for standard specific action posture.

Abstract: A data de-identification apparatus and method are provided. The data de-identification apparatus stores a data set of a first industry, wherein the data set is defined with a plurality of fields. The data de-identification apparatus receives a first instruction and a second instruction, wherein the first instruction corresponds to a second industry and the second instruction corresponds to a use of data. The data de-identification apparatus determines an identification category for each of the fields according to the first industry, the second industry, and the use of data. The data de-identification apparatus transforms the data set into a transformed data set according to the use of data and then transforms the transformed data set into a de-identification data set according to the identification categories.

Abstract: A simulation device and method are provided. The simulation device includes an interface, a storage, and a processor. The storage stores a plurality of digital twin models, each of the digital twin models simulates one entity of the at least one production line. The processor performs the following operations: generating a plurality of task requirements according to a production specification and a production data; broadcasting the task requirements to each of the digital twin models that meets one of the task requirements, wherein each of the digital twin models generates a state report based on the received task requirements; generating a plurality of task twin models based on the state reports, wherein each of the task twin models includes a group of digital twin models corresponding to the task requirements; and generating a plurality of simulation results according to the task twin models.

Abstract: A vehicle status detecting apparatus and a vehicle status detecting method thereof are provided. The vehicle status detecting apparatus stores a piece of decode-related information and a plurality of reference data, and decodes a network packet retrieved on a controller area network (CAN) to obtain a packet data according to the decode-related information. The vehicle status detecting apparatus detects pieces of electronic control unit (ECU) information which correspond to a plurality of bytes of the packet data, and transmits a warning message to the ECU when a detecting result exceeds a standard value of the plurality of reference data.

Abstract: A coach apparatus and a cooperative operation controlling method for a coach-driven multi-robot cooperative operation system are provided. The coach apparatus connects with a plurality of action robots and receives state space data from each action robot. The coach apparatus divides the action robots into a plurality of action group. The coach apparatus reduces the state space data of the action robots in each action group to generate reduced state space data of each action group, and trains and builds a cooperative model based on the reduced state space data. In addition, for each action group, the coach apparatus selects a group strategy based on the reduced state space data, and transmits the cooperative model and the group strategy to the action robots in the action group to make them carry out an action mission according to the cooperative model and the group strategy.

Abstract: A computing device divides a training image into a plurality of training blocks, and the training image includes a training object. The computing device calculates, for each of the training blocks, a correct confidence score of the training object covering the training block according to an image-marking data and a confidence-score-translating function, and the image-marking data includes a piece of location information of the training object in the training image. Then, the computing device trains a deep-learning model with the training image, the correct confidence scores and the image-marking data to generate the object-detecting model.

Abstract: A system for training transfer models that includes a memory and a processor coupling to each other. The memory stores instructions. The processor accesses the instructions to retrieve raw datasets from a first-type device and train a feature extraction model, a transfer feature and a preliminary classify model based on the raw datasets. The processor further transfers the feature extraction model, the transfer feature and the preliminary classify model to a host of a manufacturer so that the host directly applies the feature extraction model and the transfer feature, and retrains the preliminary classify model as an adapted classify model based on practical datasets.

Abstract: A speech synthesis system includes an operating interface, a storage unit and a processor. The operating interface provides a plurality of language options for a user to select one output language option therefrom. The storage unit stores a plurality of acoustic models. Each acoustic model corresponds to one of the language options and includes a plurality of phoneme labels corresponding to a specific vocal. The processor receives a text file and generates output speech data corresponding to the specific vocal according to the text file, a speech synthesizer, and one of the acoustic models which corresponds to the output language option.

Abstract: A knowledge graph generating apparatus, method and non-transitory computer readable storage medium thereof are provided. The apparatus marks an entity-relationship of the template of goods information in the template of webpage according to the operating signal and generates an extraction rule set, wherein the template of webpage is one of multiple goods webpages and has a template format. The apparatus extracts a plurality of first product information of the first goods webpages according to the extraction rule set, wherein the first goods webpages have the template format and are selected from the goods webpages. The apparatus generates a classified goods information result through a product information classification model, wherein the product information classification model is generated based on the first product information and the entity-relationship of the template of goods information.

Abstract: A method and a system for multilayer modeling are provided. The system includes a processing unit and a model building and training unit. The processing unit is configured to obtain an original data from a storage unit, obtain plural data sets of the fundamental combinations, plural data sets of the partial combinations and a data set of the full combination from the original data according to plural categorical variables of the original data, and divide the data set of each of the fundamental combinations, the data set of each of the partial combinations and the data set of the full combination into a training data set, a validation data set and a testing data set to obtain plural training data sets, plural validation data sets and plural testing data sets. The model building and training unit is configured to build plural models respectively according to the training data sets.

Abstract: An optimization method based on feature extraction and machine learning is provided. At least one input parameter is received. Multiple first historical mold data are retrieved. A similarity calculation is performed according to the input parameter and the first historical mold data. Multiple candidate mold data are selected according to the similarity calculation. The mold design parameters of the candidate mold data corresponding to each input parameter are replaced by the input parameter, and multiple first representative mold data for performing a first simulation analysis are generated. Multiple key feature parameters are found, and multiple second historical mold data are retrieved according to the multiple key feature parameters. An expected data is found, and the mold design parameters of the expected data are filtered and optimized to find multiple second representative mold data for performing a second simulation analysis. At least one set of mold production parameters is generated.

Abstract: An apparatus and method for generating a dialogue state tracking model. The apparatus retrieves a field feature corresponding to a queried field from a database according to the queried field corresponding to a queried message. The apparatus retrieves a candidate-term feature corresponding to each of at least one candidate-term corresponding to the queried field from the database, and integrates them into an integrated feature. The apparatus also generates at least one relation sub-sentence of a reply message corresponding to the queried message and generates a sentence relation feature according to the at least one relation sub-sentence. The apparatus further generates a queried field related feature according to the field feature, the integrated feature and the sentence relation feature and trains the dialogue state tracking model according to the queried field related feature.