Credits

Voices and audio samples

The French voice in the demonstration comes from a human recording in the SIWIS French Speech Synthesis Database.

Authors: Pierre-Edouard Honnet, Alexandros Lazaridis, Philip N. Garner and Junichi Yamagishi (2017). Reference excerpt: emph_siwis_s11_0034.

The four translations were prepared in advance and synthesised with the cloned voice from this excerpt. They illustrate the voice output; they are neither live translations nor a measurement of rack performance.

Uni-Translate adaptations: translation, multilingual synthesis, resampling and MP3 encoding. Licence: Creative Commons Attribution 4.0 International (CC BY 4.0).

The authors and the reference speaker do not endorse this demonstration.

Demo audio samples

The demo audio credits include MiniLibriSpeech, a subset of the LibriSpeech corpus published on OpenSLR (SLR31), under the Creative Commons Attribution 4.0 International licence (CC BY 4.0).

The original LibriSpeech corpus was prepared by Vassil Panayotov with the assistance of Daniel Povey, using audiobooks from the LibriVox project.

Globe cartography

The character-based globe uses Natural Earth land data at 1:110 million scale, version 5.1.2, adapted for this website. These data are in the public domain.

Typefaces

Syne, DM Sans and JetBrains Mono are used on the website through Google Fonts. Their respective authors distribute all three families under the SIL Open Font License 1.1.

Icons and 3D rendering

Icons come from Lucide and its contributors (ISC licence, including elements derived from Feather under MIT). 3D rendering uses Three.js and Arseny Kapoulkine’s meshoptimizer decoder, distributed under the MIT licence. Their notices are preserved with the resources.