Elhuyar has taken another step in the field of language technologies and artificial intelligence, with the aim of continuing to respond to societal challenges in the field of digitisation. On this occasion, it has created neural systems that make text voice possible and made it available to society.

Although TTS voices appear real, they are generated by a technology based on neural networks. In addition to making its own TTS voices available to society, Elhuyar offers its customers the possibility to create custom TTS voices.

Elhuyar has made available to users a range of synthetic voices that seem real, both in Basque and Spanish, and has created a reading bar that improves the accessibility of websites.

If the reading bar is inserted into a website, users will be able to listen to the contents of the website through synthetic voices. Just press the play button to listen to the content of the website and, at the same time, the part of the text that is reading the voice TTS is highlighted. The reading bar can be easily inserted on customers’ websites; by way of example, Elhuyar inserted it into these two sites: elhuyar.aldizkaria.eus and zientzia.

Elhuyar's TTS voices can be incorporated into any application through the REST API available to customers. This is how voices have been integrated into the website hiztegiak.elhuyar.eus.

Elhuyar also offers the possibility of creating custom TTS voices. That is, it offers the possibility of creating a system based on the voice of a person to whom the system imitates. Voices are personalized and unique. With this type of custom TTS voices, companies can create their own voices that allow them, among many other options, to strengthen their brand and branding strategy. For example, companies can use a personalized TTS voice in their digital communication with their customers: answering machine, audiovisual resources… In addition, these personalized voices will take care of your privacy and can only be used by the customer. More information: ttsneuronala.elhuyar.eus

Technology developed by Orai

These text-to-speech technologies are created by the artificial intelligence initiative Orai, and this new service shows the progress in research in the field of voice synthesis. Orai also addresses other lines of research, including: speech synthesis without gender marks, emotional speech synthesis and voice imitation systems from small samples. Orai is a specialized initiative in the NLP (Natural Language Processing) area, offering high-level research and smart technology solutions to companies and institutions.