Listening Dialogue Audio Attribution
=====================================

Mastermind Japanese includes pre-generated listening-comprehension dialogue
audio for pipeline validation and future learner-facing listening content.

How the audio was produced:
Dialogue audio was generated offline using the AivisSpeech Engine
(https://github.com/Aivis-Project/AivisSpeech-Engine, LGPL-3.0) with the
にせ (Nise) and まお (Mao) voice models. The engine and voice models are
licensed separately. Only the generated audio files are served from this site
as static content. We do not redistribute the AivisSpeech Engine binaries or
the voice model weights.

Voice models:

Name:          にせ (ノーマル)
AIVM page ID: 6d11c6c2-f4a4-4435-887e-23dd60f8b8dd
Speaker UUID: bf56410a-d8e6-430d-a477-f789e16206d3
Speaker ID:   1937616896
License:      ACML 1.0
Source:       https://hub.aivis-project.com/aivm-models/6d11c6c2-f4a4-4435-887e-23dd60f8b8dd

Name:          まお (ノーマル)
AIVM page ID: a59cb814-0083-4369-8542-f51a29e72af7
Speaker UUID: e756b8e4-b606-4e15-99b1-3f9c6a1b2317
Speaker ID:   888753760
License:      ACML 1.0
Source:       https://hub.aivis-project.com/aivm-models/a59cb814-0083-4369-8542-f51a29e72af7

Permissions under ACML 1.0:
ACML 1.0 permits personal, nonprofit, corporate, and commercial use of audio
generated with the voice models, subject to the prohibited-use categories listed
in the license (impersonation, defamation, misinformation, political/religious
advocacy, criminal activity, etc.). Mastermind's use — language-learning
listening comprehension and pronunciation reference audio — does not fall under
any prohibited category.

No affiliation / no endorsement:
The generated audio is not official content and is not affiliated with or
endorsed by Aivis Project, AivisHub, the model creators, or the
speakers/characters represented by the voice models. Model names are used here
solely for attribution; their use does not imply sponsorship, partnership, or
approval.

No live TTS at runtime:
No live text-to-speech service runs in the application at runtime. All listening
dialogue audio is pre-rendered and served as static MP3 files.

Current generated-audio manifest:
https://mastermindjapanese.com/legal/listening_dialogue_audio_manifest.json

For full ACML 1.0 license terms, see:
https://github.com/Aivis-Project/ACML/blob/master/ACML-1.0.md