The process of transfer a speech signal by high confidentially and as quickly as possible through the Internet needs to develop compression and encryption technology for a speech signal, so as, to reduce its size and make it understandable to persons
not authorized to listen to.
A system was designed to encrypt the voice over Internet Protocol (VoIP) and use compression technique for the purpose of reducing the size of data and send it over the network, (A_law PCM) algorithm was used the to compress audio data. Then algorithms of Triple Data Encryption Standard (TDES) and Advanced.
Encryption Standard (AES) were applied. A new encryption algorithm was proposed based in its work on the block cipher encryption system called the Direct and Reverse algorithm, which based on three basic steps, firstly expand the initial key, secondly direct the encryption of each round in one direction, and finally substitute (Bytes) as used in the Compensation Box in AES algorithm by making it moving. In general compression ratio was calculated and it was (50%) and the results of the correlation coefficient for the proposed algorithm was compared with the results of (AES, TDES) algorithms.
Personas are useful for dialogue response prediction. However, the personas used in current studies are pre-defined and hard to obtain before a conversation. To tackle this issue, we study a new task, named Speaker Persona Detection (SPD), which aims
to detect speaker personas based on the plain conversational text. In this task, a best-matched persona is searched out from candidates given the conversational text. This is a many-to-many semantic matching task because both contexts and personas in SPD are composed of multiple sentences. The long-term dependency and the dynamic redundancy among these sentences increase the difficulty of this task. We build a dataset for SPD, dubbed as Persona Match on Persona-Chat (PMPC). Furthermore, we evaluate several baseline models and propose utterance-to-profile (U2P) matching networks for this task. The U2P models operate at a fine granularity which treat both contexts and personas as sets of multiple sequences. Then, each sequence pair is scored and an interpretable overall score is obtained for a context-persona pair through aggregation. Evaluation results show that the U2P models outperform their baseline counterparts significantly.
In this paper, we use domain generalization to improve the performance of the cross-device speaker verification system. Based on a trainable speaker verification system, we use domain generalization algorithms to fine-tune the model parameters. First
, we use the VoxCeleb2 dataset to train ECAPA-TDNN as a baseline model. Then, use the CHT-TDSV dataset and the following domain generalization algorithms to fine-tune it: DANN, CDNN, Deep CORAL. Our proposed system tests 10 different scenarios in the NSYSU-TDSV dataset, including a single device and multiple devices. Finally, in the scenario of multiple devices, the best equal error rate decreased from 18.39 in the baseline to 8.84. Successfully achieved cross-device identification on the speaker verification system.
Raimy (1999; 2000a; 2000b) proposed a graphical formalism for modeling reduplication, originallymostly focused on phonological overapplication in a derivational framework. This framework is now known as Precedence-based phonology or Multiprecedence p
honology. Raimy's idea is that the segments at the input to the phonology are not totally ordered by precedence. This paper tackles a challenge that arose with Raimy's work, the development of a deterministic serialization algorithm as part of the derivation of surface forms. The Match-Extend algorithm introduced here requires fewer assumptions and sticks tighter to the attested typology. The algorithm also contains no parameter or constraint specific to individual graphs or topologies, unlike previous proposals. Match-Extend requires nothing except knowing the last added set of links.
Reliable tagging of Temporal Expressions (TEs, e.g., Book a table at L'Osteria for Sunday evening) is a central requirement for Voice Assistants (VAs). However, there is a dearth of resources and systems for the VA domain, since publicly-available te
mporal taggers are trained only on substantially different domains, such as news and clinical text. Since the cost of annotating large datasets is prohibitive, we investigate the trade-off between in-domain data and performance in DA-Time, a hybrid temporal tagger for the English VA domain which combines a neural architecture for robust TE recognition, with a parser-based TE normalizer. We find that transfer learning goes a long way even with as little as 25 in-domain sentences: DA-Time performs at the state of the art on the news domain, and substantially outperforms it on the VA domain.
Spoken language understanding (SLU) extracts the intended mean- ing from a user utterance and is a critical component of conversational virtual agents. In enterprise virtual agents (EVAs), language understanding is substantially challenging. First, t
he users are infrequent callers who are unfamiliar with the expectations of a pre-designed conversation flow. Second, the users are paying customers of an enterprise who demand a reliable, consistent and efficient user experience when resolving their issues. In this work, we describe a general and robust framework for intent and entity extraction utilizing a hybrid of statistical and rule-based approaches. Our framework includes confidence modeling that incorporates information from all components in the SLU pipeline, a critical addition for EVAs to en- sure accuracy. Our focus is on creating accurate and scalable SLU that can be deployed rapidly for a large class of EVA applications with little need for human intervention.
Paralinguistics, the non-lexical components of speech, play a crucial role in human-human interaction. Models designed to recognize paralinguistic information, particularly speech emotion and style, are difficult to train because of the limited label
ed datasets available. In this work, we present a new framework that enables a neural network to learn to extract paralinguistic attributes from speech using data that are not annotated for emotion. We assess the utility of the learned embeddings on the downstream tasks of emotion recognition and speaking style detection, demonstrating significant improvements over surface acoustic features as well as over embeddings extracted from other unsupervised approaches. Our work enables future systems to leverage the learned embedding extractor as a separate component capable of highlighting the paralinguistic components of speech.
One of the first building blocks to create a voice assistant relates to the task of tagging entities or attributes in user queries. This can be particularly challenging when entities are in the tenth of millions, as is the case of e.g. music catalogs
. Training slot tagging models at an industrial scale requires large quantities of accurately labeled user queries, which are often hard and costly to gather. On the other hand, voice assistants typically collect plenty of unlabeled queries that often remain unexploited. This paper presents a weakly-supervised methodology to label large amounts of voice query logs, enhanced with a manual filtering step. Our experimental evaluations show that slot tagging models trained on weakly-supervised data outperform models trained on hand-annotated or synthetic data, at a lower cost. Further, manual filtering of weakly-supervised data leads to a very significant reduction in Sentence Error Rate, while allowing us to drastically reduce human curation efforts from weeks to hours, with respect to hand-annotation of queries. The method is applied to successfully bootstrap a slot tagging system for a major music streaming service that currently serves several tens of thousands of daily voice queries.
Abstract We introduce Generative Spoken Language Modeling, the task of learning the acoustic and linguistic characteristics of a language from raw audio (no text, no labels), and a set of metrics to automatically evaluate the learned representations
at acoustic and linguistic levels for both encoding and generation. We set up baseline systems consisting of a discrete speech encoder (returning pseudo-text units), a generative language model (trained on pseudo- text), and a speech decoder (generating a waveform from pseudo-text) all trained without supervision and validate the proposed metrics with human evaluation. Across 3 speech encoders (CPC, wav2vec 2.0, HuBERT), we find that the number of discrete units (50, 100, or 200) matters in a task-dependent and encoder- dependent way, and that some combinations approach text-based systems.1
Voice over IP Protocol is an important Internet voice connection, characterized by high
quality of service. In this work, we will assess how today's Internet service matches its
expectations by examining the performance of the Voice over IP protoco
l and its quality of
service. We have relied on the method of selection of encoders first within some
parameters to obtain the simulation result of the comparison and analysis (QoS). Which we
adopted on VoIP protocols in the case of multiple users with three algorithms for the
symbols, and after determining the problem in this range, we took a number of factors into
account due to their impact on sound performance, such as jitter and Delay. This action
simulates three of the most common encoders (analog audio conversion and packet
compression), G.711, G.723.1 and G.729. The main objective is to achieve high-quality
sound performance by making the appropriate decision in the choice of sound encoder.
