Abstract: A system and method are presented for acoustic data selection of a particular quality for training the parameters of an acoustic model, such as a Hidden Markov Model and Gaussian Mixture Model, for example, in automatic speech recognition systems in the speech analytics field. A raw acoustic model may be trained using a given speech corpus and maximum likelihood criteria. A series of operations are performed, such as a forced Viterbi-alignment, calculations of likelihood scores, and phoneme recognition, for example, to form a subset corpus of training data. During the process, audio files of a quality that does not meet a criterion, such as poor quality audio files, may be automatically rejected from the corpus. The subset may then be used to train a new acoustic model.

Abstract: A method includes receiving a composite security identifier associated with a remote device in a local device. A plurality of identification codes associated with the remote device are encoded in the composite security identifier. An access request from the remote device is received in the local device. The access request is associated with a first one of the plurality of identification codes. The remote device is challenged for a second one of the plurality of identification codes different than the first one of the identification codes. An access level for the remote device is set on the local device based on the composite security identifier and the challenging of the remote device. The access request is selectively executed or denied based on the access level.

Abstract: A system and method are presented for acoustic data selection of a particular quality for training the parameters of an acoustic model, such as a Hidden Markov Model and Gaussian Mixture Model, for example, in automatic speech recognition systems in the speech analytics field. A raw acoustic model may be trained using a given speech corpus and maximum likelihood criteria. A series of operations are performed, such as a forced Viterbi-alignment, calculations of likelihood scores, and phoneme recognition, for example, to form a subset corpus of training data. During the process, audio files of a quality that does not meet a criterion, such as poor quality audio files, may be automatically rejected from the corpus. The subset may then be used to train a new acoustic model.

Abstract: A method includes detecting an interaction event on a display of a device in a locked mode. The interaction event originates at one of a fingerprint sensor of the device or an application icon and terminates at the other of the fingerprint sensor or the application icon. The interaction event is authenticated using the fingerprint sensor. A function associated with the application icon is activated responsive to authenticating the interaction event without unlocking the device.

Abstract: A method includes detecting an interaction event on a display of a device in a locked mode. The interaction event originates at one of a fingerprint sensor of the device or an application icon and terminates at the other of the fingerprint sensor or the application icon. The interaction event is authenticated using the fingerprint sensor. A function associated with the application icon is activated responsive to authenticating the interaction event without unlocking the device.

Abstract: An electronic device includes one or more processors and a wireless communication circuit. The one or more processors cause the wireless communication circuit to engage in a pairing procedure with at least one other electronic device to establish, during a pairing process, a peer-to-peer wireless communication link with the second electronic device. During pairing, the electronic device determines a minimum signal strength of wireless communication signals required to establish the peer-to-peer wireless communication link. Thereafter, the electronic device communicates at, and/or transmits a device discovery beacon having, a signal strength equal to or less than the minimum signal strength to conserve power.

Abstract: A method includes receiving headset motion data over a headset interface of a device while playing an audio selection on the device. A preference indicator for the audio selection is generated based on the headset motion data. A device includes a headset interface and a processor coupled to the headset interface. The processor is to receive headset motion data over the headset interface while playing an audio selection over the headset interface and generate a preference indicator for the audio selection based on the headset motion data.

Abstract: A method includes receiving an incoming biometric identification request on a first device from a user. First location data for the first device is determined. Second location data associated with at least one previous biometric identification request associated with the user is retrieved. A second level identification request is selectively initiated on the first device based on at least the first and second location data. A device includes a location module to determine first location data for the device, a biometric sensor to generate an incoming biometric identification request from a user, and a processor to retrieve second location data associated with at least one previous biometric identification request associated with the user and selectively initiate a second level identification request on the device based on at least the first and second location data.

Abstract: What is disclosed is a system and method for modelling a class posterior probability of context dependent phonemes in a speech recognition system. A representation network is trained by projecting a N-dimensional feature vector into G intermediate layers of nodes. At least some features are associated with a class label vector. A last intermediate layer ZG of the representation network is discretized to obtain a discretized layer {circumflex over (Z)}. Feature vector Q is obtained by randomly selecting V features from discretized layer {circumflex over (Z)}. Q is repeatedly hashed to obtain a vector Qf where Qf is an output of the fth hashing. An equivalent scalar representation is determined for each Qf. In a manner more fully disclosed herein, a posterior probability Pf is determined for each (x, b) pair based on the equivalent scalar representation of each respective Qf. The obtained posterior probabilities are used to improve classification accuracy in a speech recognition system.

Abstract: A method includes identifying an active state of a headset interface of a device. A signal generated by a microphone of the device is monitored to identify an alert trigger not initiated by a user of the device. Operation of the headset interface is interrupted responsive to identifying the alert trigger. A device includes a microphone, a headset interface, and a processor coupled to the microphone and the headset interface. The processor is to identify an active state of the headset interface, monitor a signal generated by the microphone to identify an alert trigger not initiated by a user of the device, and interrupt operation of the headset interface responsive to identifying the alert trigger.

Abstract: An electronic device includes a biometric sensor, such as a fingerprint sensor, that identifies biometric input received at the biometric sensor. One or more processors operable with the biometric sensor identify at least one authentication agent operating on at least one companion device operating within a wireless communication radius of the electronic device. The processors then select an authentication credential for selected authentication agents. A wireless communication circuit delivers the authentication credential to the authentication agent. An authentication tool can also be delivered.

Abstract: An electronic device includes a biometric sensor, such as a fingerprint sensor, that identifies biometric input received at the biometric sensor. One or more processors operable with the biometric sensor identify one or more companion devices operating within a wireless communication radius of the electronic device. Where multiple companion devices are within the wireless communication radius, a user can make a selection of one or more of them. One or more gesture sensors identify a predefined gesture input, such as a key turn simulation. A wireless communication circuit responsive to the one or more processors, delivers an actuation credential to at least one companion device to control the companion device.

Abstract: An electronic device includes a biometric sensor, such as a fingerprint sensor, to identify biometric input. One or more processors are then operable to identify at least one paired device and at least one companion device operating within a wireless communication radius. The paired device then detects and identifies a simple gesture, communicating this in an electronic communication to a wireless communication circuit of the electronic device. The one or more processors then identify at least one authentication agent operating on the at least one companion device and select at least one authentication credential for the at least one authentication agent. The wireless communication circuit then delivers an authentication credential in response to the predefined gesture input.

Abstract: A method and a system for detecting sentiment of a human based on an analysis of human speech are disclosed. In an embodiment, one or more time instances of glottal closure are determined from a speech signal of the human. A voice source signal based on the determined one or more time instances of glottal closure is generated. A set of relative harmonic strengths is determined based on one or more harmonic contours of the voice source signal. The RHS is indicative of a deviation of the one or more harmonics of the voice source signal from a fundamental frequency of the voice source signal. A set of feature vectors is determined based on the RHS. The set of feature vectors are utilizable to detect the sentiment of the human.

Abstract: A method, system, and computer program product for automatically receiving authorization data required to log into a service portal from a second device to a first device, via a validated connection. The method includes the first device submitting a first request to a provider to acquire authorization data required for an authenticated login to an associated online service of the provider. In response to the first device submitting the first request to the provider, the first device automatically receives the authorization data via a validated connection with the second device. In response to receiving the authorization data via the validated connection, the first device transmits a second request, including the authorization data, to the provider to gain access to the online service.

Abstract: The disclosed embodiments relate to a method of keyword recognition in a speech signal. The method includes determining a first likelihood score and a second likelihood score of one or more features of a frame of said speech signal being associated with one or more states in a first model and one or more states in a second model, respectively. The one or more states in the first model corresponds to one or more tied triphone states and the one or more states in the second model corresponds to one or more monophone states of a keyword to be recognized in the speech signal. The method further includes determining a third likelihood score based on the first likelihood score and the second likelihood score. The first likelihood score and the third likelihood score are utilizable to determine presence of the keyword in the speech signal.

Abstract: The disclosed embodiments relate to a method of keyword recognition in a speech signal. The method includes determining a first likelihood score and a second likelihood score of one or more features of a frame of said speech signal being associated with one or more states in a first model and one or more states in a second model, respectively. The one or more states in the first model corresponds to one or more tied triphone states and the one or more states in the second model corresponds to one or more monophone states of a keyword to be recognized in the speech signal. The method further includes determining a third likelihood score based on the first likelihood score and the second likelihood score. The first likelihood score and the third likelihood score are utilizable to determine presence of the keyword in the speech signal.

Abstract: A system and method for learning alternate pronunciations for speech recognition is disclosed. Alternative name pronunciations may be covered, through pronunciation learning, that have not been previously covered in a general pronunciation dictionary. In an embodiment, the detection of phone-level and syllable-level mispronunciations in words and sentences may be based on acoustic models trained by Hidden Markov Models. Mispronunciations may be detected by comparing the likelihood of the potential state of the targeting pronunciation unit with a pre-determined threshold through a series of tests. It is also within the scope of an embodiment to detect accents.

Abstract: A method and a system for detecting sentiment of a human based on an analysis of human speech are disclosed. In an embodiment, one or more time instances of glottal closure are determined from a speech signal of the human. A voice source signal based on the determined one or more time instances of glottal closure is generated. A set of relative harmonic strengths is determined based on one or more harmonic contours of the voice source signal. The RHS is indicative of a deviation of the one or more harmonics of the voice source signal from a fundamental frequency of the voice source signal. A set of feature vectors is determined based on the RHS. The set of feature vectors are utilizable to detect the sentiment of the human.

Abstract: A system and method are presented for selectively biased linear discriminant analysis in automatic speech recognition systems. Linear Discriminant Analysis (LDA) may be used to improve the discrimination between the hidden Markov model (HMM) tied-states in the acoustic feature space. The between-class and within-class covariance matrices may be biased based on the observed recognition errors of the tied-states, such as shared HMM states of the context dependent tri-phone acoustic model. The recognition errors may be obtained from a trained maximum-likelihood acoustic model utilizing the tied-states which may then be used as classes in the analysis.