ASLP-lab commited on
Commit
57a32ff
Β·
verified Β·
1 Parent(s): b34d2dd
Files changed (1) hide show
  1. README.md +44 -9
README.md CHANGED
@@ -23,13 +23,17 @@ tags:
23
  </p>
24
  <p>
25
  <a href="#introduction">Introduction</a> &nbsp;&middot;&nbsp;
 
26
  <a href="#installation">Installation</a> &nbsp;&middot;&nbsp;
27
  <a href="#inference">Inference</a> &nbsp;&middot;&nbsp;
 
28
  <a href="#citation">Citation</a>
29
  </p>
30
  </div>
31
 
32
- ## Introduction
 
 
33
 
34
  SemBridge bridges discrete semantic supervision and continuous acoustic generation.
35
  A shared semantic tokenizer anchors both acoustic latents and autoregressive language-model
@@ -41,12 +45,25 @@ The unified model supports **text-to-speech**, **note-conditioned singing**, and
41
  <img src="https://huggingface.co/ASLP-lab/SemBridge/resolve/main/asset/sembridge_model.png" alt="SemBridge architecture: semantic anchoring for continuous-latent autoregressive generation" width="100%">
42
  </div>
43
 
44
- ## News
 
 
45
 
46
  - **[2026-10]**: Released the [inference code](https://github.com/ASLP-lab/SemBridge) and [model checkpoints](https://huggingface.co/ASLP-lab/SemBridge).
47
  - **[2026-08]**: Released the [arXiv paper](https://arxiv.org/pdf/2608.07462).
48
 
49
- ## Installation
 
 
 
 
 
 
 
 
 
 
 
50
 
51
  Use **Python 3.10**.
52
 
@@ -63,7 +80,9 @@ python -m pip install torch==2.6.0 torchaudio==2.6.0 \
63
  python -m pip install -e .
64
  ```
65
 
66
- ## Inference
 
 
67
 
68
  ```python
69
  from sembridge import SemBridge
@@ -85,7 +104,9 @@ model.generate(
85
  | **SVS** | Lyrics, MIDI pitches, note durations | BigVGAN-Large |
86
  | **Unified** | Ordered speech and singing segments | BigVGAN-Large |
87
 
88
- ### Examples
 
 
89
 
90
  The repository includes **10 complete examples from our [Demo page](https://tiamojames.github.io/SemBridge_Demo/)**:
91
  4 TTS, 4 SVS, and 2 Unified cases, with reference audio and prepared requests.
@@ -110,7 +131,9 @@ Audio is saved to `outputs/examples/<example_id>.wav`. Set `MODEL=/path/to/SemBr
110
  to use local weights. See the **[example catalog](https://github.com/ASLP-lab/SemBridge/blob/main/examples/README.md)** for the
111
  complete list, reference recordings, and guidance on using your own songs.
112
 
113
- ### Single-Sample Inference
 
 
114
 
115
  ```bash
116
  # Text-to-speech
@@ -142,14 +165,24 @@ For SVS, each note contains `text`, MIDI `pitch`, and `duration` **in seconds**.
142
  See the **[inference guide](https://github.com/ASLP-lab/SemBridge/blob/main/docs/INFERENCE.md)** for reference voices, singing and
143
  mixed-generation examples, sampling controls, and offline loading.
144
 
145
- ## Acknowledgements
 
 
 
 
 
 
 
 
146
 
147
  We thank the authors of the following projects for their open research and resources:
148
 
149
  - **Speech generation:** [VoxCPM](https://github.com/OpenBMB/VoxCPM) and [DiTAR](https://arxiv.org/abs/2502.03930).
150
  - **Audio codecs:** [DAC](https://github.com/descriptinc/descript-audio-codec), [SAC](https://github.com/Soul-AILab/SAC), and [BigVGAN](https://github.com/NVIDIA/BigVGAN).
151
 
152
- ## Citation
 
 
153
 
154
  If you find SemBridge useful in your research, please cite:
155
 
@@ -166,7 +199,9 @@ If you find SemBridge useful in your research, please cite:
166
  }
167
  ```
168
 
169
- ## License
 
 
170
 
171
  SemBridge code is released under Apache-2.0; upstream portions retain their
172
  original licenses. See [LICENSE](https://github.com/ASLP-lab/SemBridge/blob/main/LICENSE) for terms and third-party notices.
 
23
  </p>
24
  <p>
25
  <a href="#introduction">Introduction</a> &nbsp;&middot;&nbsp;
26
+ <a href="#roadmap">Roadmap</a> &nbsp;&middot;&nbsp;
27
  <a href="#installation">Installation</a> &nbsp;&middot;&nbsp;
28
  <a href="#inference">Inference</a> &nbsp;&middot;&nbsp;
29
+ <a href="#community">Community</a> &nbsp;&middot;&nbsp;
30
  <a href="#citation">Citation</a>
31
  </p>
32
  </div>
33
 
34
+ <a id="introduction"></a>
35
+
36
+ ## πŸ“– Introduction
37
 
38
  SemBridge bridges discrete semantic supervision and continuous acoustic generation.
39
  A shared semantic tokenizer anchors both acoustic latents and autoregressive language-model
 
45
  <img src="https://huggingface.co/ASLP-lab/SemBridge/resolve/main/asset/sembridge_model.png" alt="SemBridge architecture: semantic anchoring for continuous-latent autoregressive generation" width="100%">
46
  </div>
47
 
48
+ <a id="news"></a>
49
+
50
+ ## πŸ“£ News
51
 
52
  - **[2026-10]**: Released the [inference code](https://github.com/ASLP-lab/SemBridge) and [model checkpoints](https://huggingface.co/ASLP-lab/SemBridge).
53
  - **[2026-08]**: Released the [arXiv paper](https://arxiv.org/pdf/2608.07462).
54
 
55
+ <a id="roadmap"></a>
56
+
57
+ ## πŸ—ΊοΈ Roadmap
58
+
59
+ - [x] [arXiv paper](https://arxiv.org/abs/2608.07462) and [code repository](https://github.com/ASLP-lab/SemBridge)
60
+ - [x] [TTS, SVS, and unified inference code](#inference)
61
+ - [x] [Model checkpoints](https://huggingface.co/ASLP-lab/SemBridge)
62
+ - [ ] Fine-tuning code
63
+
64
+ <a id="installation"></a>
65
+
66
+ ## βš™οΈ Installation
67
 
68
  Use **Python 3.10**.
69
 
 
80
  python -m pip install -e .
81
  ```
82
 
83
+ <a id="inference"></a>
84
+
85
+ ## πŸš€ Inference
86
 
87
  ```python
88
  from sembridge import SemBridge
 
104
  | **SVS** | Lyrics, MIDI pitches, note durations | BigVGAN-Large |
105
  | **Unified** | Ordered speech and singing segments | BigVGAN-Large |
106
 
107
+ <a id="examples"></a>
108
+
109
+ ### 🎧 Examples
110
 
111
  The repository includes **10 complete examples from our [Demo page](https://tiamojames.github.io/SemBridge_Demo/)**:
112
  4 TTS, 4 SVS, and 2 Unified cases, with reference audio and prepared requests.
 
131
  to use local weights. See the **[example catalog](https://github.com/ASLP-lab/SemBridge/blob/main/examples/README.md)** for the
132
  complete list, reference recordings, and guidance on using your own songs.
133
 
134
+ <a id="single-sample-inference"></a>
135
+
136
+ ### πŸŽ™οΈ Single-Sample Inference
137
 
138
  ```bash
139
  # Text-to-speech
 
165
  See the **[inference guide](https://github.com/ASLP-lab/SemBridge/blob/main/docs/INFERENCE.md)** for reference voices, singing and
166
  mixed-generation examples, sampling controls, and offline loading.
167
 
168
+ <a id="community"></a>
169
+
170
+ ## 🀝 Community
171
+
172
+ - **Windows / ComfyUI:** [ComfyUI-SemBridge](https://github.com/hiroki-abe-58/ComfyUI-SemBridge), an unofficial, community-maintained integration.
173
+
174
+ <a id="acknowledgements"></a>
175
+
176
+ ## πŸ™ Acknowledgements
177
 
178
  We thank the authors of the following projects for their open research and resources:
179
 
180
  - **Speech generation:** [VoxCPM](https://github.com/OpenBMB/VoxCPM) and [DiTAR](https://arxiv.org/abs/2502.03930).
181
  - **Audio codecs:** [DAC](https://github.com/descriptinc/descript-audio-codec), [SAC](https://github.com/Soul-AILab/SAC), and [BigVGAN](https://github.com/NVIDIA/BigVGAN).
182
 
183
+ <a id="citation"></a>
184
+
185
+ ## πŸ“š Citation
186
 
187
  If you find SemBridge useful in your research, please cite:
188
 
 
199
  }
200
  ```
201
 
202
+ <a id="license"></a>
203
+
204
+ ## πŸ“„ License
205
 
206
  SemBridge code is released under Apache-2.0; upstream portions retain their
207
  original licenses. See [LICENSE](https://github.com/ASLP-lab/SemBridge/blob/main/LICENSE) for terms and third-party notices.