Abstract: The present disclosure discloses an image processing device including: a receiving module configured to receive a voice signal and an image to be processed; a conversion module configured to convert the voice signal into an image processing instruction and determine a target area according to a target voice instruction conversion model, in which the target area is a processing area of the image to be processed; and a processing module configured to process the target area according to the image processing instruction and a target image processing model. The examples may realize the functionality of using voice commands to control image processing, which may save users' time spent in learning image processing software prior to image processing, and improve user experience.

Abstract: A device and method for improving hearing devices by using computer recognition of words and substituting either computer generated words or pre-recorded words in streaming conversation received from a distant speaker. The system may operate in multiple modes such as a first mode being amplification and conditioning of the voice sounds; a second mode having said microphone pickup up the voice sounds from a speaker, a processor configured to convert voice sounds to discrete words corresponding to words spoken by said speaker, generating a synthesized voice speaking said words and outputting said synthesized voice to said sound reproducing element, which is hearable by the user. Other modes include translation of foreign languages into a user's ear and using a heads up display to project the text version of words which the computer had deciphered or translated. The system may be triggered by eye moment, spoken command, hand movement or similar.

Abstract: A method may include generating, based a context-free grammar, a sample forming a corpus. The context-free grammar may include production rules for replacing a first nonterminal symbol with a second nonterminal symbol and/or a terminal symbol. The sample may be generated by rewriting recursively a first text string to form a second text string associated with the sample. The first text string may be rewritten by applying the production rules to replace nonterminal symbols included in the first text string until no nonterminal symbols remain in the first text string. A machine learning model may be trained, based on the corpus, to process a natural language. Related methods and articles of manufacture are also disclosed.

Abstract: A device and a method for clarifying dysarthria voices is disclosed. Firstly, a dysarthria voice signal is received and framed to generate dysarthria frames. Then, the dysarthria frames are received to retrieve dysarthria features. Finally, the dysarthria features are received. Without receiving phases corresponding to the dysarthria features, the dysarthria features are converted into an intelligent voice signal based on an intelligent voice conversion model. The intelligent voice conversion model is not trained by the dynamic time warping (DTW). The present invention avoids the phase distortion of the voice signal and provides more natural and clarified voices with low noise.

Abstract: Methods and systems are provided for a natural language processing system comprising a chatbot adapted for dialog generation. In one example, the system may include a combination of a variational autoencoder (VAE) and a generative adversarial network (GAN) for generating natural responses to input queries. The VAE may convert queries into vector embeddings that may then be used by the GAN to continuously update and improve responses provided by the chatbot.

Abstract: A speech translation device, for conversation between a first speaker making an utterance in a first language and a second speaker making an utterance in a second language different from the first language, includes: a speech detector that detects, from sounds that are input, a speech segment in which the first speaker or the second speaker made an utterance; a display that, after speech recognition is performed on the utterance, displays a translation result obtained by translating the utterance from the first language to the second language or from the second language to the first language; and an utterance instructor that outputs, in the second language via the display, a message prompting the second speaker to make an utterance after a first speaker's utterance or outputs, in the first language via the display, a message prompting the first speaker to make an utterance after a second speaker's utterance.

Abstract: A method includes receiving, from an electronic device, information defining a user utterance associated with a skill to be performed, where the skill is not recognized by a natural language understanding (NLU) engine. The method also includes receiving, from the electronic device, information defining one or more actions for performing the skill. The method further includes identifying, using at least one processor, one or more known skills having one or more slots that map to at least one word or phrase in the user utterance. The method also includes creating, using the at least one processor, a plurality of additional utterances based on the one or more mapped slots. In addition, the method includes training, using the at least one processor, the NLU engine using the plurality of additional utterances.

Abstract: It is an aspect of the present disclosure to provide a dialogue system capable of providing an extended function to the user by registering a new vocabulary that matches the user's preference and by changing the pre-stored conversation pattern.

Abstract: Disclosed is a regional-features-based speech recognition method, including learning speech features by region using speech data classified by region category, and recognizing input speech using an acoustic model and a language model generated through classification of a region category for the input speech and the learning. A user may use a dialect recognition service that is improved using learning based on artificial intelligence (AI) and enhanced mobile broadband (eMBB), ultra-reliable and low latency communications (URLLC), and massive machine-type communications (mMTC) techniques of 5G mobile communication.

Abstract: A new approach to automatic speech recognition is disclosed. An example method include receiving a first text representing speech recognition of a phrase spoken by a user, isolating a candidate named entity from within the phrase, receiving a first phonetic representation of the candidate named entity, comparing the first phonetic representation to phonetic representations in a mapping database which map the phonetic representations to words to yield a comparison, based on the comparison, identifying a second phonetic representation in the mapping database that matches a second text in the mapping database to the second phonetic representation and replacing the candidate named entity with the second text. The approach can be used for new brands for which automatic speech recognition error rates are high.

Abstract: Disclosed are an end-to-end system for speech recognition and speech translation and an electronic device. The system comprises an acoustic encoder and a multi-task decoder and a semantic invariance constraint module, and completes two tasks for speech recognition and speech translation. In addition, according to the characteristic of the semantic consistency of texts between different tasks, semantic constraints are imposed on the model to learn high-level semantic information, and the semantic information can effectively improve the performance of speech recognition and speech translation. The application has the following advantages that the error accumulation problem of serial system is avoided, and the calculation cost of the model is low and the real-time performance is very high.

Abstract: The present disclosure discloses an image processing device including: a receiving module configured to receive a voice signal and an image to be processed; a conversion module configured to convert the voice signal into an image processing instruction and determine a target area according to a target voice instruction conversion model, in which the target area is a processing area of the image to be processed; and a processing module configured to process the target area according to the image processing instruction and a target image processing model. The examples may realize the functionality of using voice commands to control image processing, which may save users' time spent in learning image processing software prior to image processing, and improve user experience.

Abstract: System, methods, and computer readable media can be used to create a virtual assistant. One of the methods includes receiving audio from a conversation between two parties while the conversation is occurring. The method includes generating a partial transcript of the conversation. The method includes identifying topics based on the partial transcript. The method includes presenting a user interface element based on the identified topic.

Abstract: The present disclosure discloses an image processing device including: a receiving module configured to receive a voice signal and an image to be processed; a conversion module configured to convert the voice signal into an image processing instruction and determine a target area according to a target voice instruction conversion model, in which the target area is a processing area of the image to be processed; and a processing module configured to process the target area according to the image processing instruction and a target image processing model. The examples may realize the functionality of using voice commands to control image processing, which may save users' time spent in learning image processing software prior to image processing, and improve user experience.

Abstract: Method for providing assistance in conversation including recognizing, by recognition module, conversation between primary user and at least one secondary user, identifying, by recognition module, first and second context data for primary user and at least one secondary user based on conversation; generating, by response generation module, at least one response on behalf of primary user based on at least one of second context data derived from at least one secondary user, and first context data; analyzing, by determining module, at least one action of primary user in at least one response on second context data; determining, by determining module, intervening situation in conversation based on at least one action; selecting, by intervening response module, intervening response from at least one response for determined intervening situation based on at least one action; and delivering, by response delivery module, intervening response to at least one secondary user during determined intervening situation.

Abstract: The present invention proposes a new method for improving the performance of a real-valued filterbank based spectral envelope adjuster. By adaptively locking the gain values for adjacent channels dependent on the sign of the channels, as defined in the application, reduced aliasing is achieved. Furthermore, the grouping of the channels during gain-calculation, gives an improved energy estimate of the real valued subband signals in the filterbank.

Abstract: Systems and methods are described herein for generating an adaptive response to a user request. Input indicative of a user request may be received and utilized to identify an item in an electronic catalog. Title segments may be identified from the item's title. Significant segments of the user request may be determined. In response to the user request, a shortened title may be generated from the identified title segments and provided as output at the user device (e.g., via audible output provided at a speaker of the user device, via textual output, or the like). At least one of the title segments provided in the shortened title may correlate to the significant segment identified from the user request. In some embodiments, the length and content of the shortened title may vary based at least in part on the contextual intent of the user's request.

Abstract: An electronic apparatus is provided. The electronic device includes: a storage configured to store recognition related information and misrecognition related information of a trigger word for entering a speech recognition mode; and a processor configured to identify whether or not the speech recognition mode is activated on the basis of characteristic information of a received uttered speech and the recognition related information, identify a similarity between text information of the received uttered speech and text information of the trigger word, and update at least one of the recognition related information or the misrecognition related information on the basis of whether or not the speech recognition mode is activated and the similarity.

Abstract: Methods, systems, and apparatus, including computer programs stored on a computer-readable storage medium, for transliteration for speech recognition training and scoring. In some implementations, language examples are accessed, some of which include words in a first script and words in one or more other scripts. At least portions of some of the language examples are transliterated to the first script to generate a training data set. A language model is generated based on occurrences of the different sequences of words in the training data set in the first script. The language model is used to perform speech recognition for an utterance.

Abstract: A method for detecting an audio signal and an apparatus, where the method includes determining a segmental signal-to-noise ratio (SSNR) of an audio signal in response to the audio signal being an unvoiced signal, reducing a reference voice activity detection (VAD) decision threshold to obtain a reduced VAD decision threshold, and comparing the SSNR with the reduced VAD decision threshold to determine whether the audio signal is an active signal.