Abstract: Technologies pertaining to training a deep neural network (DNN) for use in a recognition system are described herein. The DNN is trained using heterogeneous data, the heterogeneous data including narrowband signals and wideband signals. The DNN, subsequent to being trained, receives an input signal that can be either a wideband signal or narrowband signal. The DNN estimates the class posterior probability of the input signal regardless of whether the input signal is the wideband signal or the narrowband signal.

Abstract: Systems and methods for speech recognition incorporating environmental variables are provided. The systems and methods capture speech to be recognized. The speech is then recognized utilizing a variable component deep neural network (DNN). The variable component DNN processes the captured speech by incorporating an environment variable. The environment variable may be any variable that is dependent on environmental conditions or the relation of the user, the client device, and the environment. For example, the environment variable may be based on noise of the environment and represented as a signal-to-noise ratio. The variable component DNN may incorporate the environment variable in different ways. For instance, the environment variable may be incorporated into weighting matrices and biases of the DNN, the outputs of the hidden layers of the DNN, or the activation functions of the nodes of the DNN.

Abstract: A centrifugal fan for an electronic device is described. The centrifugal fan includes a volute casing with an outlet; and vanes accommodated in the volute casing. The volute casing has an inner surface facing the vanes, the inner surface being formed with a plurality of air guiding channels for directing air inside the volute casing in a direction towards the outlet of the volute casing.

Abstract: Systems and methods for identifying a false representation of a human face are provided. In one example, a method for identifying a false representation of a human face includes receiving a plurality of different data streams captured by a respective plurality of sensors of differing sensor types sensing a candidate face. In a cascading plurality of stages, one or more of the different data streams are analyzed, wherein each of the stages comprises a different analysis. In one of the cascading plurality of stages, the method determines that one or more of the different data streams corresponds to a false representation of the human face. Based on determining that one or more of the different data streams corresponds to a false representation of a human face, an indication of the false representation is outputted.

Abstract: A high-dimensional posterior-based feature with partial distance elimination may be utilized for speech recognition. The log likelihood values of a large number of Gaussians are needed to generate the high-dimensional posterior feature. Gaussians with very small log likelihoods are associated with zero posterior values. Log likelihoods for Gaussians for a speech frame may be evaluated with a partial distance elimination method. If the partial distance of a Gaussian is already too small, the Gaussian will have a zero posterior value. The partial distance may be calculated by sequentially adding individual dimensions in a group of dimensions. The partial distance elimination occurs when less than all of the dimensions in the group are sequentially added.

Abstract: The adaptation and personalization of a deep neural network (DNN) model for automatic speech recognition is provided. An utterance which includes speech features for one or more speakers may be received in ASR tasks such as voice search or short message dictation. A decomposition approach may then be applied to an original matrix in the DNN model. In response to applying the decomposition approach, the original matrix may be converted into multiple new matrices which are smaller than the original matrix. A square matrix may then be added to the new matrices. Speaker-specific parameters may then be stored in the square matrix. The DNN model may then be adapted by updating the square matrix. This process may be applied to all of a number of original matrices in the DNN model. The adapted DNN model may include a reduced number of parameters than those received in the original DNN model.

Abstract: Systems and methods are provided for generating a DNN classifier by “learning” a “student” DNN model from a larger more accurate “teacher” DNN model. The student DNN may be trained from un-labeled training data because its supervised signal is obtained by passing the un-labeled training data through the teacher DNN. In one embodiment, an iterative process is applied to train the student DNN by minimize the divergence of the output distributions from the teacher and student DNN models. For each iteration until convergence, the difference in the output distributions is used to update the student DNN model, and output distributions are determined again, using the unlabeled training data. The resulting trained student model may be suitable for providing accurate signal processing applications on devices having limited computational or storage resources such as mobile or wearable devices. In an embodiment, the teacher DNN model comprises an ensemble of DNN models.

Abstract: Techniques and systems for training an acoustic model are described. In an embodiment, a technique for training an acoustic model includes dividing a corpus of training data that includes transcription errors into N parts, and on each part, decoding an utterance with an incremental acoustic model and an incremental language model to produce a decoded transcription. The technique may further include inserting silence between a pair of words into the decoded transcription and aligning an original transcription corresponding to the utterance with the decoded transcription according to time for each part. The technique may further include selecting a segment from the utterance having at least Q contiguous matching aligned words, and training the incremental acoustic model with the selected segment. The trained incremental acoustic model may then be used on a subsequent part of the training data. Other embodiments are described and claimed.

Abstract: A video game system (or other data processing system) can visually identify a person entering a field of view of the system and determine whether the person has been previously interacting with the system. In one embodiment, the system establishes thresholds, enrolls players, performs the video game (or other application) including interacting with a subset of the players based on the enrolling, determines that a person has become detectable in the field of view of the system, automatically determines whether the person is one of the enrolled players, maps the person to an enrolled player and interacts with the person based on the mapping if it is determined that the person is one of the enrolled players, and assigns a new identification to the person and interacts with the person based on the new identification if it is determined that the person is not one of the enrolled players.

Abstract: Systems and methods for identifying a false representation of a human face are provided. In one example, a method for identifying a false representation of a human face includes receiving a plurality of different data streams captured by a respective plurality of sensors of differing sensor types sensing a candidate face. In a cascading plurality of stages, one or more of the different data streams are analyzed, wherein each of the stages comprises a different analysis. In one of the cascading plurality of stages, the method determines that one or more of the different data streams corresponds to a false representation of the human face. Based on determining that one or more of the different data streams corresponds to a false representation of a human face, an indication of the false representation is outputted.

Abstract: Systems and methods for identifying a false representation of a human face are provided. In one example, a method for identifying a false representation of a human face includes receiving a plurality of different data streams captured by a respective plurality of sensors of differing sensor types sensing a candidate face. In a cascading plurality of stages, one or more of the different data streams are analyzed, wherein each of the stages comprises a different analysis. In one of the cascading plurality of stages, the method determines that one or more of the different data streams corresponds to a false representation of the human face. Based on determining that one or more of the different data streams corresponds to a false representation of a human face, an indication of the false representation is outputted.

Abstract: The present invention discloses an airflow accelerating device and an electronic apparatus. The airflow accelerating device comprises: a housing, having a chamber formed therein; at least one vibrating plate, disposed within the chamber; at least one division plate, fixed in the housing, for dividing the chamber into at least two sub-chambers, each of the at least two sub-chambers having at least one air outlet configured to transmit airflow generated by vibration of the vibrating plate to outside of the chamber.

Abstract: According to this disclosure, a shift register unit includes a pull-up control module, a pull-up module, a pull-down control module and a pull-down module, wherein the pull-up module is adapted to provide a transmission signal output terminal with a first clock signal inputted from a first clock signal input terminal according to the pull-up control signal, and provide a gate drive signal output terminal with a first direct current supply voltage according to the pull-up control signal and the first clock signal inputted from the first clock signal input terminal.

Abstract: Providing a framework for merging automatic speech recognition (ASR) systems having a shared deep neural network (DNN) feature transformation is provided. A received utterance may be evaluated to generate a DNN-derived feature from the top hidden layer of a DNN. The top hidden layer output may then be utilized to generate a network including a bottleneck layer and an output layer. Weights representing a feature dimension reduction may then be extracted between the top hidden layer and the bottleneck layer. Scores may then be generated and combined to merge the ASR systems which share the DNN feature transformation.

Abstract: Systems and methods for face recognition are provided. In one example, a method for face recognition includes receiving a user image and detecting a user luminance of data representing the user's face. An adaptive low pass filter is selected that corresponds to the user luminance of the user's face. The filter is applied to the user image to create a filtered user image. The filtered user image is projected to create a filtered user image representation. A filtered reference image representation that has been filtered with the same low pass filter is selected from a reference image database. The method then determines whether the filtered reference image representation matches the filtered user image representation.

Abstract: The adaptation and personalization of a deep neural network (DNN) model for automatic speech recognition is provided. An utterance which includes speech features for one or more speakers may be received in ASR tasks such as voice search or short message dictation. A decomposition approach may then be applied to an original matrix in the DNN model. In response to applying the decomposition approach, the original matrix may be converted into multiple new matrices which are smaller than the original matrix. A square matrix may then be added to the new matrices. Speaker-specific parameters may then be stored in the square matrix. The DNN model may then be adapted by updating the square matrix. This process may be applied to all of a number of original matrices in the DNN model. The adapted DNN model may include a reduced number of parameters than those received in the original DNN model.

Abstract: A heat dissipating device located on a same base plate with a fan, and a first side of the heat dissipating device abutting an air outlet of the fan. The heat dissipating device includes a plurality of horizontal heat dissipating fins horizontally arranged in parallel with each other in a stacked manner such that arrangement of the plurality of horizontal heat dissipating fins is consistent with a direction of the air outlet of the fan, wherein two adjacent horizontal heat dissipating fins have a gap there between, through which air from the air outlet of the fan is discharged, wherein, a first horizontal heat dissipating fin located at an uppermost layer is coupled with a heat pipe and the heat pipe is coupled with at least one heating element. The heat dissipating device is capable of enhancing heat dissipating efficiency and improving the user's usage experience.

Abstract: Described is a calibration model for use in a speech recognition system. The calibration model adjusts the confidence scores output by a speech recognition engine to thereby provide an improved calibrated confidence score for use by an application. The calibration model is one that has been trained for a specific usage scenario, e.g., for that application, based upon a calibration training set obtained from a previous similar/corresponding usage scenario or scenarios. Different calibration models may be used with different usage scenarios, e.g., during different conditions. The calibration model may comprise a maximum entropy classifier with distribution constraints, trained with continuous raw confidence scores and multi-valued word tokens, and/or other distributions and extracted features.

Abstract: A computing system such as a game console maintains and updates a biometric profile of a user. In one aspect, biometric data of the user is continuously obtained from a sensor such as an infrared and visible light camera, and used to update the biometric profile using a machine learning process. In another aspect, a user is prompted to confirm his or her identify when multiple users are detected at the same time and/or when the user is detected with a confidence level which is below a threshold. A real-time image of the user being identified can be displayed on a user interface with user images associated with one or more accounts. In another aspect, the biometric profile is managed by a shell on the computing system, where the shell makes the biometric profile available to any of a number of applications on the computing system.

Abstract: A video game system (or other data processing system) can visually identify a person entering a field of view of the system and determine whether the person has been previously interacting with the system. In one embodiment, the system establishes thresholds, enrolls players, performs the video game (or other application) including interacting with a subset of the players based on the enrolling, determines that a person has become detectable in the field of view of the system, automatically determines whether the person is one of the enrolled players, maps the person to an enrolled player and interacts with the person based on the mapping if it is determined that the person is one of the enrolled players, and assigns a new identification to the person and interacts with the person based on the new identification if it is determined that the person is not one of the enrolled players.