KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#dsh #dsh-plugin #ai-agents #developer-tools #codex #claude-code #mcp #rust #deepseek-harness #typescript #ai-agent #cli

CED

★ 76 Open GitHub ↗

Source code for Consistent ensemble distillation for audio tagging

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Consistent Ensemble Distillation for Audio Tagging (CED)

This repo is the source for the ICASSP 2024 Paper [Consistent Ensemble Distillation for Audio Tagging](https://arxiv.org/abs/2308.11957).

![Framework](https://raw.githubusercontent.com/RicherMans/CED/main/figures/plot_performance.png)

| Model | Parameters (M) | AS-20K (mAP) | AS-2M (mAP) |
|------|-------|-------|-------|
| CED-Tiny | 5.5   | 36.5  | 48.1  |
| CED-Mini | 9.6    | 38.5  | 49.0  |
| CED-Small| 22    | 41.6  | 49.6  |
| CED-Base | 86    | 44.0  | 50.0  |

* All models work with 16 kHz audio and use 64-dim Mel-spectrograms, making them very fast. `CED-Tiny` should be faster than MobileNets on a single x86 CPU (even though MACs/FLops would indicate otherwise).

Pretrained models could be downloaded from [Zenodo](https://zenodo.org/record/8275347) or [Hugging Face](https://huggingface.co/mispeech).

|  | Zenodo | Hugging Face |
|------|:-------:|:-------:|
| CED-Tiny | [Link](https://zenodo.org/record/8275347/files/audiotransformer_tiny_mAP_4814.pt?download=1) | [Link](https://huggingface.co/mispeech/ced-tiny) |
| CED-Mini | [Link](https://zenodo.org/record/8275347/files/audiotransformer_mini_mAP_4896.pt?download=1) | [Link](https://huggingface.co/mispeech/ced-mini) |
| CED-Small| [Link](https://zenodo.org/record/8275319/files/audiotransformer_small_mAP_4958.pt?downloa)   | [Link](https://huggingface.co/mispeech/ced-small)|
| CED-Base | [Link](https://zenodo.org/record/8275347/files/audiotransformer_base_mAP_4999.pt?download=1) | [Link](https://huggingface.co/mispeech/ced-base) |

## Demo

We have an online demo available [here](https://huggingface.co/spaces/mispeech/ced-base) for CED-Base.

## Inference/Usage

### Huggingface Transformers

#### Inference (Onnx, Recommended)

```python 
>>> from optimum.onnxruntime import ORTModelForAudioClassification

>>> model_name = "mispeech/ced-mini"
>>> model = ORTModelForAudioClassification.from_pretrained(model_name, trust_remote_code=True)

>>> import torchaudio
>>> audio, sampling_rate = torchaudio.load("/path-to/JeD5V5aaaoI_931_932.wav")
>>> assert sampling_rate == 16000
>>> input_name = model.session.get_inputs()[0].name
>>> output = model(**{input_name: torch.randn(1, 16000)})
>>> logits = output.logits.squeeze()
>>> for idx in logits.argsort()[-2:][::-1]:
>>>   print(f"{model.config.id2label[idx]}: {logits[idx]:.4f}")
'Finger snapping: 0.9155'
'Slap: 0.0567'
```

Or standard Huggingface Transformers

```python
# pip install transformers
from transformers import AutoModelForAudioClassification, AutoFeatureExtractor

model_name = "mispeech/ced-base"
feature_extractor = AutoFeatureExtractor.from_pretrained(model_name, trust_remote_code=True)
model = AutoModelForAudioClassification.from_pretrained(model_name, trust_remote_code=True)

import torchaudio
audio, sampling_rate = torchaudio.load("/path-to/JeD5V5aaaoI_931_932.wav")
assert sampling_rate == 16000
inputs = feature_extractor(audio, sampling_rate=sampling_rate, return_tensors="pt")

import torch
with torch.no_grad():
    logits = model(**inputs).logits

predicted_class_id = torch.argmax(logits, dim=-1).item()
model.config.id2label[predicted_class_id]
```

### Locally

To just use the CED models for inference, simply run:

```bash
git clone https://github.com/Richermans/CED/
cd CED/
pip3 install -r requirements.txt
python3 inference.py resources/*
```

Note that I experienced some problems with higher versions of `hdf5`, so if possible please use `1.12.1`.

By default we use `CED-mini` here, which offers a good trade-off between performance and speed.
One can switch the models with the -m flag:

```bash
python3 inference.py -m ced_tiny resources/*
python3 inference.py -m ced_mini resources/*
python3 inference.py -m ced_small resources/*
python3 inference.py -m ced_base resources/*

```

You can also use the models directly from Hugging Face, see [here](https://huggingface.co/mispeech/ced-base) for usage instructions.

## Training/Reproducing results

### 1. Preparing data

First, one needs to download Audioset. One might use one of our own [scripts](https://github.com/RicherMans/SAT/blob/main/datasets/audioset/1_download_audioset.sh).

For example, one can put the downloaded files into a folder named `data/balanced` and `data/unbalanced`, `data/eval` such as:

```
data/balanced/
├── -0DdlOuIFUI_50.000.wav
├── -0DLPzsiXXE_30.000.wav
├── -0FHUc78Gqo_30.000.wav
├── -0mjrMposBM_80.000.wav
├── -0O3e95y4gE_100.000.wav
…

data/unbalanced/
├── --04kMEQOAs_0.000_10.000.wav
├── --0aJtOMp2M_30.000_40.000.wav
├── --0AzKXCHj8_22.000_32.000.wav
├── --0B3G_C3qc_10.000_20.000.wav
├── --0bntG9i7E_30.000_40.000.wav
…

data/eval/
├── 007P6bFgRCU_10.000_20.000.wav
├── 00AGIhlv-w0_300.000_310.000.wav
├── 00FBAdjlF4g_30.000_40.000.wav
├── 00G2vNrTnCc_10.000_20.000.wav
├── 00KM53yZi2A_30.000_40.000.wav
├── 00XaUxjGuX8_170.000_180.000.wav
├── 0-2Onbywljo_380.000_390.000.wav

```

Then just generate a .tsv file with:

```
find data/balanced/ -type f | awk 'BEGIN{print "filename"}{print}' > data/balanced.tsv
```

Then dump the data as `hdf5` files using `scripts/wavlist_to_hdf5.py`:

```bash
python3 scripts/wavlist_to_hdf5.py data/balanced.tsv data/balanced_train/
```

This will generate a training datafile `data/balanced_train/labels/balanced.tsv`.

For the eval data, please use [this script](https://github.com/RicherMans/SAT/blob/main/datasets/audioset/1_download_audioset.sh) to download.

The resulting `eval.tsv` should look like this:

```
filename	labels	hdf5path
data/eval/--4gqARaEJE.wav	73;361;74;72	data/eval_data/hdf5/eval_0.h5
data/eval/--BfvyPmVMo.wav	419	data/eval_data/hdf5/eval_0.h5
data/eval/--U7joUcTCo.wav	47	data/eval_data/hdf5/eval_0.h5
data/eval/-0BIyqJj9ZU.wav	21;20;17	data/eval_data/hdf5/eval_0.h5
data/eval/-0Gj8-vB1q4.wav	273;268;137	data/eval_data/hdf5/eval_0.h5
data/eval/-0RWZT-miFs.wav	379;307	data/eval_data/hdf5/eval_0.h5
data/eval/-0YUDn-1yII.wav	268;137	data/eval_data/hdf5/eval_0.h5
data/eval/-0jeONf82dE.wav	87;137;89;0;72	data/eval_data/hdf5/eval_0.h5
data/eval/-0nqfRcnAYE.wav	364	data/eval_data/hdf5/eval_0.h5
```

### 2. Download logits

Download the logits used in the paper from [Zenodo](https://zenodo.org/record/8275347/files/logits.zip?download=1):

```bash
wget https://zenodo.org/record/8275347/files/logits.zip?download=1 -O logits.zip
unzip logits.zip
```

This will create:

```bash
logits/
└── ensemble5014
    ├── balanced
    │   └── chunk_10
    └── full
        └── chunk_10
```

### 3. Train

```bash
python3 run.py train trainconfig/balanced_mixup_tiny_T_ensemble5014_chunk10.yaml
```

## Export ONNX and Inference ONNX

```bash
python3 export_onnx.py -m ced_tiny
#or ced_mini ced_small ced_base
```

```bash
python3 onnx_inference_with_kaldi.py test.wav -m ced_tiny.onnx
python3 onnx_inference_with_torchaudio.py test.wav -m ced_tiny.onnx
```
Why use Kaldi to calculate Mel features? Because it has ready-made C++ implementation code, which can be found here: https://github.com/csukuangfj/kaldi-native-fbank/tree/master

## Training on your own data

This is a label-free framework, meaning that any data can be used for optimization.
To use your own data, do the follwing:

Put your data somewhere and generate a `.tsv` file with a single header `filename`, such as:

```bash
find some_directory -type f | awk 'BEGIN{print "filename"}{print}' > my_data.tsv
```

Then dump the corresponding hdf5 file using `scripts/wavlist_to_hdf5.py`:

```bash
python3 scripts/wavlist_to_hdf5.py my_data.tsv my_data_hdf5/
```

Then run the script `save_logits.py` as:

```bash
torchrun save_logits.py logitconfig/balanced_base_chunk10s_topk20.yaml --train_data my_data_hdf5/labels/my_data.tsv
```

Finally you can train your own model on that augmented dataset with:

```bash
python3 run.py train trainconfig/balanced_mixup_base_T_ensemble5014_chunk10.yaml --logitspath YOUR_LOGITS_PATH --train_data YOUR_TRAIN_DATA.tsv
```

### Hear-Evaluation

We also submitted the models for the HEAR benchmark evaluation.
Hear uses a simple linear downstream evaluation protocol across 19 tasks. 
We simply extracted the features from all ced-models from the penultimate layer.
The repo can be found [here](https://github.com/RicherMans/HEAR_CED).

| Model                                      | Beehive States Avg | Beijing Opera Percussion | CREMA-D | DCASE16 | ESC-50 | FSD50K | GTZAN Genre | GTZAN Music Speech | Gunshot Triangulation | LibriCount | MAESTRO 5hr | Mridangam Stroke | Mridangam Tonic | NSynth Pitch 50hr | NSynth Pitch 5hr | Speech Commands 5hr | Speech Commands Full | Vocal Imitations | VoxLingua107 Top10 |
|--------------------------------------------|--------------------|--------------------------|---------|---------|--------|--------|-------------|-------------------|----------------------|------------|-------------|------------------|------------------|-------------------|-----------------|--------------------|---------------------|-----------------|--------------------|
| ced-tiny | 38.345             | 94.90                    | 62.52   | 88.02   | 95.80  | 62.73  | 89.20       | 93.01             | 91.67                | 61.26      | 4.81        | 96.13            | 90.74            | 69.19             | 44.00           | 70.53              | 77.10               | 19.18           | 33.64              |
| ced-mini     | 59.17              | 96.18                    | 65.26   | 90.66   | 95.35  | 63.88  | 90.30       | 94.49             | 86.01                | 64.02      | 8.29        | 96.56            | 93.32            | 75.20             | 55.60           | 77.38              | 81.96               | 20.37           | 34.67              |
| ced-small     | 51.70              | 96.60                    | 66.64   | 91.63   | 95.95  | 64.33  | 89.50       | 91.22             | 93.45                | 65.59      | 10.96       | 96.82            | 93.94            | 79.95             | 60.20           | 80.92              | 85.19               | 21.92           | 36.53              |
| ced-base      | 48.35              | 96.60                    | 69.10   | 92.19   | 96.65  | 65.48  | 88.60       | 94.36             | 89.29                | 67.85      | 14.76       | 97.43            | 96.55            | 82.81             | 68.20           | 86.93              | 89.67               | 22.69           | 38.57              |

### Android APks

Thanks to [csukuangfj](https://github.com/csukuangfj), there are also pre-compiled android binaries using [sherpa-onnx](https://github.com/k2-fsa/sherpa-onnx).

Binaries are available on the k2-fsa sherpa page, https://k2-fsa.github.io/sherpa/onnx/audio-tagging/apk.html.

### Citation

Please cite our paper if you find this work useful:

```bibtex
@inproceedings{dinkel2023ced,
  title={CED: Consistent ensemble distillation for audio tagging},
  author={Dinkel, Heinrich and Wang, Yongqing and Yan, Zhiyong and Zhang, Junbo and Wang, Yujun},
  booktitle={ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  year={2024}
}
```

CED

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README