| karin v1 — Japanese vowel-frame classifier (ONNX) |
| Copyright (c) 2026 Spectopathy |
|
|
| Licensed under the Apache License, Version 2.0 (the "License"). |
|
|
| This model was trained using the following voice resources. Distributions of |
| this model (including derivative works that embed this model) must reproduce |
| this NOTICE per Apache License 2.0, Section 4(d). |
|
|
| -------------------------------------------------------------------------------- |
|
|
| Teacher acoustic backbone |
| reazon-research/japanese-wav2vec2-base-rs35kh |
| © Human Interaction Laboratory |
| https://huggingface.co/reazon-research/japanese-wav2vec2-base-rs35kh |
| License: Apache License 2.0 |
|
|
| ReazonSpeech (small subset) — distillation audio |
| © Human Interaction Laboratory |
| https://huggingface.co/datasets/reazon-research/reazonspeech |
| https://research.reazon.jp/ |
| License: CDLA-Sharing 1.0 |
|
|
| つくよみちゃんコーパス Vol.1 |
| © 夢前黎 |
| 利用規約: https://tyc.rei-yumesaki.net/material/corpus/ |
|
|
| あみたろのITAコーパス読み上げ音声 |
| © あみたろの声素材工房 |
| 利用規約: https://amitaro.net/voice/corpus-list/ita/ |
|
|
| AivisHub 公開モデルによる合成音声（複数） |
| https://hub.aivis-project.com/ |
| License: ACML 1.0 / CC0 |
|
|
