245results about How to "Efficient learning process" patented technology

A method and system for logical group endpoint discovery in a data communication network is disclosed. A network endpoint receives from a source a request identifying a logical group and determines whether a port associated with the endpoint is a member of the logical group. If a port associated with the endpoint is a member of the logical group, the endpoint transmits to the source a reply identifying the port and the logical group. The request may be multicast while the reply may be unicast. The logical group may be a VLAN or a multicast group. The source and endpoint are preferably 802.1 ag maintenance endpoints. The request may include an IEEE 802.1 ag continuity check multicast address as a destination address.

The present invention discloses a visual target detection and labeling method. The method includes: an image inputting step, to input an image to be detected; a candidate region extracting step, to extract a candidate window as the candidate region from the image to be detected using selectively search algorithm; a feature description extracting step, to perform feature description on the candidate region using a pre-trained large-scale convolutional neural network and output the feature description of the candidate region; a visual target predicting step, to predict the candidate region based on the feature description of the candidate region using a pre-trained object detection module, to estimate regions having the visual target; and a position labeling step, to labeling the position of the visual target according to the estimated result. Experiments show that, compared with the mainstream week supervision visual target detection and labeling method, the present invention has a stronger ability to excavate positive samples and a more general application prospect, and is suitable for visual target detection and automatic labeling tasks on the large-scale data set.

Recognition results associated with handwritten electronic ink, voice recognition results or other forms of user input can be corrected by designating at least a portion of a visual display. Recognition results corresponding to the designated portion, and optionally, additional recognition results to provide context, are displayed. Portions of the displayed recognition results are then selected, and alternate recognition results made available. Alternate recognition results can be chosen, and the selected recognition results modified based upon the chosen alternate. The invention may include different user interfaces.

Jigsaw connector pieces representing tasks to be performed are displayed on a computer screen, preferably hierarchically in a tree, to elucidate the relationships between tasks. To perform a task, a user positions an attachment jigsaw piece next to its corresponding connector piece. The user is then prompted to input or verify preferences in a properties panel. For example, a printer may be installed by interconnecting an attachment piece (corresponding to the printer to be installed) next to the printer connector piece in the tree. The user may be aided in the process of matching connector and attachment pieces through the use of matching colors and images on the pieces.

A personal computing device comprising: a processor, an onboard memory, an accelerometer, a gyroscope, and a display; a computer program to create an exercise analysis application comprising: a software module configured to receive data from the accelerometer and the gyroscope that are associated with the bodily motion of a user in three dimensions; a software module configured to place the device in a learning mode, the learning mode comprising recording the data of the user performing a defined exercise to generate a statistical model for the exercise; a software module configured to place the device in a normal mode, the normal mode comprising applying a probabilistic analysis to the bodily motion data to identify an exercise event, classify the exercise by comparison to a recorded model; and a software module configured to apply an analysis to the bodily motion data to score the user's exercise form.

A large-scale real-time traffic flow prediction method of the present invention is based on fuzzy logic and deep LSTM, which relates to a technical field of urban intelligent traffic management. The method includes steps of: selecting an urban road network scene to collect color images of real-time traffic flow congestion information; obtaining congestion levels of multiple intersections according to the color images, which are used in a data training set; and forming a data sensing end of FDFP through a fuzzy mechanism; establishing a deep LSTM neural network, performing deep learning on the training data set, and constructing a prediction end of the FDFP; construct a graph of road intersections and formulate a k-nearest neighbors-based discounted averaging for obtaining congestion on the edges; and inputting real-time traffic information received from a server into an FDFP model.

An object image detecting apparatus for determining whether an object image exists in a detection target image includes: an image reading unit reading a predetermined region of the detection target image as a detection target region; a feature vector producing unit dividing an image included in the detection target region resized to a predetermined size into a plurality of blocks, calculating a representative value of an image feature amount indicating a predetermined image feature for each of the divided blocks, and producing a feature vector showing a characteristic of the image feature within the detection target region by referring to the representative value; and at least two classifying units determining whether the object image exists in the detection target region by different references based on the image feature indicated by the feature vector.

The invention discloses an adaptive cruise control method based on approximate policy iteration. The adaptive cruise control method comprises the steps of (1), collecting samples; (2), learning on the samples by using an approximate policy iteration algorithm to obtain an approximately optimal policy; (3), optimizing PI controller parameters online in the cruise control, namely, optimizing the PI controller parameters online in a data driving mode by using the approximately optimal policy, so that the cruise control can achieve expected performances. The adaptive cruise control method has the advantages that the principle is simple, online optimization can be achieved, control performances can be improved, and the like.

A voxel feature learning network receives a raw point cloud and converts the point cloud into a sparse 4D tensor comprising three-dimensional coordinates (e.g. X, Y, and Z) for each voxel of a plurality of voxels and a fourth voxel feature dimension for each non-empty voxel. In some embodiments, convolutional mid layers further transform the 4D tensor into a high-dimensional volumetric representation of the point cloud. In some embodiments, a region proposal network identifies 3D bounding boxes of objects in the point cloud based on the high-dimensional volumetric representation. In some embodiments, the feature learning network and the region proposal network are trained end-to-end using training data comprising known ground truth bounding boxes, without requiring human intervention.

The present invention provides a method of classifying defects wherein defects are detected in a first inspection machine. The detected defects are then reviewed by a second inspection machine. A sampling rate for review by the second inspection machine is determined by a defect classifier in the first inspection machine.

The present invention relates, in general, to a karaoke system, and, more particularly, to a karaoke system having a song learning function that enables a user to repeatedly listen to songs on a bar or length basis and enables the user to sing songs with accompaniment sounds. The present invention provides a system and method that enables the complete or bar-based singer's song to be repeatedly played back in response to a user's request, thereby enabling the user to sufficiently and conveniently practice one or more bard difficult to sing. The present invention provides a system and method that enables bar-based scores to be indicated, so that the user can be aware of one or more incorrect bars and can intensively practice the corresponding portions using the above-described function, thereby increasing the user's interest and enabling efficient learning.

The invention provides a play assist device which allows a player to efficiently learn how to read and play a score. The device includes display means for displaying the upper and lower staves of a grand staff independently by a predetermined range, projecting means for projecting the images of the upper and lower staves of a grand staff above or onto the keyboard, pressed key detection means for detecting the pressed key, and control means for shifting the upper and lower staves so that the note to be played in the upper and lower staves is positioned above or on the key corresponding to the note.

An integrated teaching base station, especially at least two interactive digital display teaching base stations located at different teaching and training sites, the one within is a first personal digital display base system, the other one at a remote terminal is the second personal digital display base system. Via broadband internet connection, teaching materials can be transmitted among the interactive digital display teaching base stations and can be accessed and displayed at any time. The teaching base stations can be further combined with an audio/video surveillance module, an audio/video data integration module and a network audio/video data communication module to form the interactive digital display audio/video teaching base station that can be used to record and store teaching data at the first personal digital data base system. Multimedia mobile communication device can be used to connect the interactive digital display audio/video teaching base station so as to observe the class at the teaching and training sites (S) in order to enhance understanding and communication. In addition, it can be used to establish a terrarium module, an aquatic module, a zoology module, etc, at the teaching and training site (S). This together with the interactive digital display audio/video teaching base station forms an interactive digital display audio/video lab environment teaching base station. A remote second personal digital data base system connects with multiple labs, operation room for clinical teaching, etc. In this way, open and detailed observations and discussions can be conducted to promote the best results in observation, research and teaching.

Provided is a speech recognition method for a recognition target language. According to an embodiment of the inventive concept, a speech recognition method for a recognition target language performed by a speech recognition apparatus includes obtaining an original learning data set for the recognition target language, constructing a target label by dividing the text information included in each piece of original learning data in letter units, and building an acoustic model based on a deep neural network by learning the learning speech data included in the each piece of original learning data and the target label corresponding to the learning speech data.

Systems and methods are provided that enable young children (e.g., toddlers to pre-kindergarten aged children) and/or persons with disabilities to easily play, and effectively learn from, interactive, educational, and entertaining games. The edutainment game may be a series of instructions stored on a digital storage medium for use with a game system. The young child or person with a disability may play the game using a wireless, handheld controller that is easily graspable, maneuverable, and manipulatable, and the young child or person with a disability may play the game alone or with supervision from another person. The edutainment games may include one or more learning tools (or learning games), emphasizing skills such as, for example, color, shape, and/or sound recognition; puzzle-solving; navigation; number recognition, counting, and/or ordering; alphabet and reading/writing; etc. These skills may be emphasized individually or in various combinations.

A method and apparatus to obtain unique study content that is selected based on user's study goals and unique preferences includes having a user input a study goal and study preference. Then a first database of learning content, that may be traditional learning content prepared by an education institution, is queried based on the input study goal and preferences and study items are extracted from the first database. Then, the study items extracted from the first database are stored in a second database, the study items stored in the second database are then used to query a third database of potential learning content, which may include non-traditional learning content, and by matching the study items stored in the second database to content in the third database, study content that has been specifically tailored and personalized based on the user's input study goals and preferences is extracted and presented to the user for study and learning. Since the proven learning content from the first database of learning items is used to match up and extract learning content from a third database of possibly non-traditional learning content, the user is assured to obtain the input study goal using content that is new, flexibly and adaptively obtained and personalized to each user or student, whenever the student wants to obtain fresh, new or different content.

The present invention relates to a method, apparatus and system for learning plan analysis. According to one embodiment of the present invention, the method for learning plan analysis includes: providing a learner with learning materials commensurate with a learning progress or learning ability; receiving learning information responsive to the learning materials; and generating analysis data from analyzing interaction patterns of a user in learning.

A method for computer-aided control of any technical system is provided. The method includes two steps, the learning of the dynamic with historical data based on a recurrent neural network and a subsequent learning of an optimal regulation by coupling the recurrent neural network to a further neural network. The recurrent neural network has a hidden layer comprising a first and a second hidden state at a respective time point. The first hidden state is coupled to the second hidden state using a matrix to be learned. This allows a bottleneck structure to be created, in that the dimension of the first hidden state is smaller than the dimension of the second hidden state or vice versa. The autonomous dynamic is taken into account during the learning of the network, thereby improving the approximation capacity of the network. The technical system includes a gas turbine.

An exercise system, apparatus, and method. The system and apparatus include at least one prompt physically associated with a user, wherein each prompt is directly corresponded to a target area of the user, and at least one instruction, wherein each instruction is associated with at least one prompt, and wherein taking action on the prompt in accordance with the instruction impacts the target area. The method includes physically associating at least one prompt with an exerciser, directly corresponding each of said at least one prompt is to a target on the exerciser, and allowing for instruction to the exerciser, wherein the instruction is associated with at least one prompt, and wherein an action on the prompt in accordance with the instruction impacts the target.

A method and system of generating, rating, and storing a pronunciation corpus is provided. The system (“Dico”) is an interactive system resident on a data network such as the Internet or intranet. Dico provides a platform for maintaining and serving the corpus in such a way that the corpus can be expanded continuously with new phrases and new pronunciations received from the users of Dico. A user of Dico can take the role of a contributor or a listener. Contributors use Dico's contribution tool to contribute new pronunciations and phrases to Dico's corpus. Listeners use Dico's playback tool to listen to the contributed pronunciations in Dico's corpus. Listeners can also rate the contributed pronunciations using Dico's rating tool. Dico uses the ratings to determine the quality of the contributed pronunciations and use this information to rank the pronunciations. The collective actions and knowledge of Dico's users enable Dico to determine the best pronunciations for each phrase in its corpus.

A machine learning method that may reduce an annotation cost and may improve performance of a target model is provided. Some embodiments of the present disclosure may provide a machine learning method performed by a computing device, including: acquiring a training dataset of a first model including a plurality of data samples to which label information is not given; calculating a miss-prediction probability of the first model on the plurality of data samples; configuring a first data sample group by selecting at least one data sample from the plurality of data samples based on the calculated miss-prediction probability; acquiring first label information on the first data sample group; and performing first learning on the first model by using the first data sample group and the first label information.

A method and apparatus for performing speech interaction with children is provided. The apparatus may be a computing device that includes at least one camera, at least one microphone, memory, and at least one processor for executing stored instructions. The at least one processor may be configured to determine an age range or an age or skill level of the child. The computing device may receive one or more inputs from the child. The at least one processor may perform analysis on the one or more inputs based at least in part on the determined age range or the age or the skill level of the child, and output a speech response to the child based on the performed analysis.