Editing audio narration using conventional software typically involves many painstaking low-level manipulations. Some state of the art systems allow the editor to work in a text transcript of the narration, and perform select, cut, copy and paste operations directly in the transcript; these operations are then automatically applied to the waveform in a straightforward manner. However, an obvious gap in the text-based interface is the ability to type new words not appearing in the transcript, for example inserting a new word for emphasis or replacing a misspoken word. While high-quality voice synthesizers exist today, the challenge is to synthesize the new word in a voice that matches the rest of the narration. This paper presents a system that can synthesize a new word or short phrase such that it blends seamlessly in the context of the existing narration. Our approach is to use a text to speech synthesizer to say the word in a generic voice, and then use voice conversion to convert it into a voice that matches the narration. Offering a range of degrees of control to the editor, our interface supports fully automatic synthesis, selection among a candidate set of alternative pronunciations, fine control over edit placements and pitch profiles, and even guidance by the editors own voice. The paper presents studies showing that the output of our method is preferred over baseline methods and often indistinguishable from the original voice.

VoCo: text-based insertion and replacement in audio narration

In this paper, we present a new objective prediction model for synthetic speech naturalness. It can be used to evaluate Text-To-Speech or Voice Conversion systems and works language independently. The model is trained end-to-end and based on a CNN-LSTM network that previously showed to give good results for speech quality estimation. We trained and tested the model on 16 different datasets, such as from the Blizzard Challenge and the Voice Conversion Challenge. Further, we show that the reliability of deep learning-based naturalness prediction can be improved by transfer learning from speech quality prediction models that are trained on objective POLQA scores. The proposed model is made publicly available and can, for example, be used to evaluate different TTS system configurations.

Deep Learning Based Assessment of Synthetic Speech Naturalness.

In recent years, rapid advances in speech technology have been made possible by machine learning challenges such as CHiME, REVERB, Blizzard, and Hurricane. In the Clarity project, the machine learning approach is applied to the problem of hearing aid processing of speech-in-noise, where current technology in enhancing the speech signal for the hearing aid wearer is often ineffective. The scenario is a (simulated) cuboid-shaped living room in which there is a single listener, a single target speaker and a single interferer, which is either a competing talker or domestic noise. All sources are static, the target is always within ±30◦ azimuth of the listener and at the same elevation, and the interferer is an omnidirectional point source at the same elevation. The target speech comes from an open source 40- speaker British English speech database collected for this purpose. This paper provides a baseline description of the round one Clarity challenges for both enhancement (CEC1) and prediction (CPC1). To the authors’ knowledge, these are the first machine learning challenges to consider the problem of hearing aid speech signal processing

Clarity-2021 challenges : machine learning challenges for advancing hearing aid processing

The world of statistical parametric speech synthesis continues to improve with recent investigations of different machine learning techniques to better model spectrum, F0 and duration from corpora of natural speech. Traditional techniques rely on decision trees alone. This paper shows the advantages of modeling with random forests of decision trees over single trees. Improvements equivalent to more than doubling the data can be achieved, offering end users significantly better synthesis from the same data size. These techniques give proportionally more improvements on smaller datasets, particularly with voices with only 30 minutes of speech. These techniques have been tested over a wide range of voices and languages of various sizes and quality, producing significant improvements in all cases. These techniques are documented, and robustly implemented for others to use through the Dec 2014 release of the Festvox voice building toolkit, thereby directly allowing these benefits to be used in standard voices build for the Festival Speech Synthesis System and CMU Flite.

Random forests for statistical speech synthesis.

Building text-to-speech (TTS) synthesisers is a difficult task, especially for low resource languages. Language-specific modules need to be developed for system building. End-to-end speech synthesis has become a popular paradigm as a TTS can be trained using only  pairs. However, end-to-end speech synthesis is not scalable in a multilanguage scenario, as the vocabulary increases with the number of different scripts. In this paper, TTSes are trained for Indian languages using two text representations– character-based and phone-based. For the character-based approach, a multi-language character map (MLCM) is proposed to easily train Indic speech synthesisers. The phone-based approach uses the common label set (CLS) representation for Indian languages. Both approaches leverage the similarities that exist among the languages. The advantage is a compact representation across multiple languages. Experiments are conducted by building TTSes using monolingual data and by pooling data across two languages. The ability to synthesise code-mixed text using the phone-based approach is also assessed. Subjective evaluations indicate that reasonably good quality Indic TTSes can be developed using both approaches. This emphasises the need to incorporate multilingual text processing in the end-to-end framework.

Building Multilingual End-to-End Speech Synthesisers for Indian Languages

In this paper, we discuss a consortium effort on building text to speech (TTS) systems for 13 Indian languages. There are about 1652 Indian languages. A unified framework is therefore attempted required for building TTSes for Indian languages. As Indian languages are syllable-timed, a syllable-based framework is developed. As quality of speech synthesis is of paramount interest, unit-selection synthesizers are built. Building TTS systems for low-resource languages requires that the data be carefully collected an annotated as the database has to be built from the scratch. Various criteria have to addressed while building the database, namely, speaker selection, pronunciation variation, optimal text selection, handling of out of vocabulary words and so on. The various characteristics of the voice that affect speech synthesis quality are first analysed. Next the design of the corpus of each of the Indian languages is tabulated. The collected data is labeled at the syllable level using a semiautomatic labeling tool. Text to speech synthesizers are built for all the 13 languages, namely, Hindi, Tamil, Marathi, Bengali, Malayalam, Telugu, Kannada, Gujarati, Rajasthani, Assamese, Manipuri, Odia and Bodo using the same common framework. The TTS systems are evaluated using degradation Mean Opinion Score (DMOS) and Word Error Rate (WER). An average DMOS score of ≈3.0 and an average WER of about 20 % is observed across all the languages.

A syllable-based framework for unit selection synthesis in 13 Indian languages

The Blizzard challenge 2014 was the tenth annual Blizzard challenge organized by the following group of institutions : IIIT Hyderabad, IIT Madras, DAIICT, SSN College of Engineering, IIT Mandi and IIT Guwahati with support and collaboration from DeitY, Government of India. This paper describes the tasks in the Blizzard challenge 2014. The tasks consisted of data from six Indian languages : Assamese, Gujarati, Hindi, Rajasthani, Tamil and Telugu. Seven participants from around the world used the speech data provided as well as the corresponding text transcriptions in UTF-8, to build synthetic voices, which were then evaluated by means of listening tests. Index Terms: Blizzard challenge, Speech synthesis, Evaluation of synthetic speech

/pdf/the-blizzard-challenge-2014-3r26fofivr.pdf

The Blizzard Challenge 2014

Indian agriculture has made rapid progress on the agricultural front during the past three decades and is in a queue of the major producer in the world. But still it has long way to go and meet challenges ahead such as communication, resources, and availabi lity at right time at right place. Theweb has had an amazing existence and it has been the driving force for a cause to grow information across boundaries, enabling effective communication and 24x7 service availability all leading to a digital information based economy that we have today. Despite that, its direct influence has reached to a small percentage of human popul ation. Since localization populated withIndia and the applications are translated and adapted for Indian u sers. With the possible localization of spread raw formatted Indian government data, at different locations are thought to have integrated with each other using the internet web technology as  Semantic Web Network.

/pdf/semantic-web-query-on-e-governance-data-and-designing-3pmcxn9lt6.pdf

Semantic web query on e- governance data and designing ontology for agriculture domain

Ecological assessment and suitability ranges of Ban oak (Quercus oblongata D. Don) in Chamba district, Himalayas: implications for present and future conservation

Maize is the main staple food of the world but it is nutritional deficient due to the scarcity of two essential amino acids viz., lysine and tryptophan. Under this study, opaque2 gene from QPM donor (CML-169) was transferred into normal maize (BAJIM-08-26) through marker aided backcross breeding method. The opaque2 specific SSR marker ( phi057 ) was used for foreground selection in BC 1 F 1 and BC 2 F 1 generations. The heterozygous offspring were self-pollinated to produce the BC 2 F 2 generation. Further plants on the basis of 25% opaqueness were selected and selfed to develop BC 2 F 3 and BC 2 F 4 progenies. The background selection using a series of SSR markers showed 97% recovery in the recurrent parent genome of backcrossed generation. Tryptophan content and total protein in the endosperm of BC 2 F 4 progenies were found to be ranged from 0.75 to 0.93% and 7.0 to 9.6 %, respectively. As an outcome, three promising MAS derived QPM lines viz., B76-22-2, B70-5-8 and B42-3-1 were developed that showed a high percentage of tryptophan, grain yield and recovery of recurrent parent genome. The worldwide problem of protein malnutrition could be improved by adopting a molecular breeding approach for the development of nutritionally rich maize over the normal maize.

Swaran Lata

Papers

A syllable-based framework for unit selection synthesis in 13 Indian languages

The Blizzard Challenge 2014

Semantic web query on e- governance data and designing ontology for agriculture domain

Ecological assessment and suitability ranges of Ban oak (Quercus oblongata D. Don) in Chamba district, Himalayas: implications for present and future conservation

Development of tryptophan rich maize lines through simple sequence repeat marker aided introgression of opaque2 trait