README.md

---
license: apache-2.0
library_name: lingvo
tags:
  - audio
  - speech-enhancement
  - echo-cancellation
  - textual-echo-cancellation
  - tensorflow
  - tflite
datasets:
  - libritts
  - vctk
metrics:
  - wer
  - mcd
arxiv: 2008.06006
---

# Vanilla-Seq2seq Baseline (No Side Input) — Multiple Interfering Voices (`NoSideInputMultiInterfering`)

Single-source sequence-to-sequence speech enhancement baseline (`Vanilla-Seq2seq` / `NoSideInput`) trained on 24 kHz **LibriTTS** user speech mixed at 0 dB SNR with reverberant **CSTR VCTK** without any side input.

- **GitHub Repository**: [https://github.com/wq2012/tec](https://github.com/wq2012/tec)
- **PyPI Package**: [`textual-echo-cancellation`](https://pypi.org/project/textual-echo-cancellation/)
- **Paper**: [Textual Echo Cancellation (IEEE SLT 2021, arXiv:2008.06006v4)](https://arxiv.org/pdf/2008.06006)
- **Audio Demo Page**: [https://google.github.io/speaker-id/publications/TEC/](https://google.github.io/speaker-id/publications/TEC/)

> **Open-Source Reproduction Notice**: This pretrained model was trained using the standalone open-source reproduction library ([`wq2012/tec`](https://github.com/wq2012/tec)) built on [`lingvo`](https://github.com/tensorflow/lingvo) and `tensorflow`. It does not use Google's proprietary internal codebase or internal training data infrastructure.

---

## Model Performance

### 1. Open-Source Reproduction Evaluation (`NoSideInputMultiInterfering`)

Evaluated on 24 kHz **LibriTTS** (`test-clean`, `test-other`) mixed at 0 dB SNR with synthetic room impulse responses (RT60 = 0.25 s) under the **Multiple interfering voices (LibriTTS + VCTK)** condition, using `Qwen3-ASR-0.6B-F16` via `audio.cpp` for ASR WER scoring and 13-MFCC Dynamic Time Warping for MCD:

| Metric | `test-clean` | `test-other` |Paper Reference (`Vanilla-Seq2seq`) |
| :--- | :---: | :---: | :---: |
| **WER (%) ↓** | **31.16%** (62/199) | **42.39%** (103/243) | 19.7% / 38.7% |
| **MCD (dB) ↓** | **7.93 dB** | **8.72 dB** | 7.53 dB / 8.87 dB |
| **Side Input Size (KB) ↓** | **0.000 KB** | **0.000 KB** | 0 KB |
| **Computational Complexity ↓** | **6.32 GFLOPS** | **6.32 GFLOPS** | 6.32 GFLOPS |

### 2. Full Comparison Across All Pretrained Models & Baselines

| Condition | Method | Hugging Face Model | WER (%) test-clean ↓ | WER (%) test-other ↓ | MCD (dB) test-clean ↓ | MCD (dB) test-other ↓ | Side Input test-clean (KB) ↓ |
| :--- | :--- | :--- | :---: | :---: | :---: | :---: | :---: |
| **Single interfering voice** *(LibriTTS + LJSpeech)* | `GroundTruth` | — | 3.52 | 6.78 | 0.00 | 0.00 | 0.000 |
| | `MicrophoneSignal` | — | 90.45 | 114.12 | 12.86 | 14.61 | 0.000 |
| | `NlmsAec` (AEC-NLMS) | — | 88.44 | 107.34 | 12.80 | 14.48 | 243.465 |
| | `NoSideInputSingleInterfering` | [`wq2012/vanilla_seq2seq_single_interfering`](https://huggingface.co/wq2012/vanilla_seq2seq_single_interfering) | 45.23 | 91.53 | 9.58 | 11.34 | 0.000 |
| | `AecSingleInterfering` | [`wq2012/aec_single_interfering`](https://huggingface.co/wq2012/aec_single_interfering) | 12.06 | 23.16 | 8.85 | 9.86 | 243.465 |
| | **`TecSingleInterfering`** | **[`wq2012/tec_single_interfering`](https://huggingface.co/wq2012/tec_single_interfering)** | **21.61** | **46.89** | **8.24** | **9.28** | **0.076** |
| **Multiple interfering voices** *(LibriTTS + VCTK)* | `GroundTruth` | — | 5.03 | 7.82 | 0.00 | 0.00 | 0.000 |
| | `MicrophoneSignal` | — | 34.17 | 48.97 | 7.67 | 7.70 | 0.000 |
| | `NlmsAec` (AEC-NLMS) | — | 28.64 | 34.98 | 7.92 | 8.40 | 186.922 |
| | `NoSideInputMultiInterfering` | [`wq2012/vanilla_seq2seq_multi_interfering`](https://huggingface.co/wq2012/vanilla_seq2seq_multi_interfering) | 31.16 | 42.39 | 7.93 | 8.72 | 0.000 |
| | `AecMultiInterfering` | [`wq2012/aec_multi_interfering`](https://huggingface.co/wq2012/aec_multi_interfering) | 8.54 | 22.22 | 7.80 | 7.88 | 186.922 |
| | **`TecMultiInterfering`** | **[`wq2012/tec_multi_interfering`](https://huggingface.co/wq2012/tec_multi_interfering)** | **26.63** | **45.27** | **7.96** | **8.40** | **0.037** |

---

## Files in This Repository

- `best.ckpt.data-00000-of-00001`, `best.ckpt.index`, `best.ckpt.meta`, `checkpoint`: TensorFlow / Lingvo checkpoint for `NoSideInputMultiInterfering`.
- `model.tflite`: Dynamic-range quantized TensorFlow Lite (`.tflite`) FlatBuffer model for on-device inference.
- `evaluation_metrics.json`: Verified evaluation results (`test-clean` and `test-other`) for `NoSideInputMultiInterfering`.

---

## How to Use

### 1. Install `textual-echo-cancellation`

```bash
pip3 install textual-echo-cancellation huggingface_hub
```

### 2. Download the Model from Hugging Face

```python
from huggingface_hub import snapshot_download

model_dir = snapshot_download(repo_id="wq2012/vanilla_seq2seq_multi_interfering")
print("Downloaded model to:", model_dir)
```

### 3. Run Inference via CLI (`scripts/inference.py`)

```bash
python3 -m scripts.inference \
  --model NoSideInputMultiInterfering \
  --checkpoint_path "${MODEL_DIR}/best.ckpt" \
  --mixed_wav /path/to/mixed_input.wav \
  # No side input required for Vanilla-Seq2seq \
  --output_wav /tmp/enhanced_clean.wav
```

### 4. Run Inference via Python API

```python
import os
from huggingface_hub import snapshot_download
from tec import inference

model_dir = snapshot_download(repo_id="wq2012/vanilla_seq2seq_multi_interfering")
ckpt_path = os.path.join(model_dir, "best.ckpt")

result = inference.run_inference_on_wav(
    model_name="NoSideInputMultiInterfering",
    mixed_wav_path="/path/to/mixed_input.wav",
    interfering_text="currently in mountain view it is 72 degrees",
    checkpoint_path=ckpt_path,
    output_wav_path="/tmp/enhanced_clean.wav",
)
print("Enhanced log-Mel spectrogram shape:", result["predicted_mel"].shape)
```

### 5. On-Device Inference with Quantized TFLite (`model.tflite`)

```python
import os
import numpy as np
import tensorflow as tf
from huggingface_hub import snapshot_download

model_dir = snapshot_download(repo_id="wq2012/vanilla_seq2seq_multi_interfering")
tflite_path = os.path.join(model_dir, "model.tflite")

interpreter = tf.lite.Interpreter(model_path=tflite_path)
interpreter.allocate_tensors()

for detail in interpreter.get_input_details():
    interpreter.set_tensor(
        detail["index"], np.zeros(detail["shape"], dtype=detail["dtype"])
    )

interpreter.invoke()
output_details = interpreter.get_output_details()
enhanced_mel = interpreter.get_tensor(output_details[0]["index"])
print("TFLite predicted log-Mel shape:", enhanced_mel.shape)
```

---

## Citation

If you use this model or the `textual-echo-cancellation` library in your research, please cite the original paper:

```bibtex
@inproceedings{ding2021textual,
  title={Textual Echo Cancellation},
  author={Ding, Shaojin and Jia, Ye and Hu, Ke and Wang, Quan},
  booktitle={2021 IEEE Spoken Language Technology Workshop (SLT)},
  pages={666--673},
  year={2021},
  organization={IEEE}
}
```