|
--- |
|
license: cc-by-nc-sa-4.0 |
|
--- |
|
|
|
### Description |
|
|
|
This model is used to separate reverb and delay effects in vocals. In addition, it can also separate partial harmony, but it cannot completely separate them. I added random high cut after the reverberation and delay effects in the dataset, so the model's handling of high frequencies is not particularly aggressive.<br> |
|
You can try listening to the performance of this model [here](https://huggingface.co/Sucial/Dereverb-Echo_Mel_Band_Roformer/tree/main/examples)! |
|
|
|
### How to use the model? |
|
|
|
Try it with [ZFTurbo's Music-Source-Separation-Training](https://github.com/ZFTurbo/Music-Source-Separation-Training) |
|
|
|
### Model |
|
|
|
### V2 Models |
|
|
|
Finetuned from: `dereverb-echo_mel_band_roformer_sdr_10.0169.ckpt`<br> |
|
Used 1000+ songs to Finetune. |
|
|
|
Config: [config_dereverb_echo_mbr_v2.yaml](./config_dereverb_echo_mbr_v2.yaml)<br> |
|
Model: [dereverb_echo_mbr_v2_sdr_dry_13.4843.ckpt](./dereverb_echo_mbr_v2_sdr_dry_13.4843.ckpt)<br> |
|
Instr dry sdr: 13.4843 (Std: 4.8675) |
|
|
|
#### V1 Models |
|
|
|
Configs_256_8_4: [config_dereverb-echo_mel_band_roformer.yaml](./config_dereverb-echo_mel_band_roformer.yaml)<br> |
|
Model_256_8_4: [dereverb-echo_mel_band_roformer_sdr_10.0169.ckpt](./dereverb-echo_mel_band_roformer_sdr_10.0169.ckpt)<br> |
|
Instr dry sdr: 13.1507, Instr other sdr: 6.8830, Metric avg sdr: 10.0169 |
|
|
|
Configs_128_4_4: [config_dereverb-echo_128_4_4_mel_band_roformer.yaml](./config_dereverb-echo_128_4_4_mel_band_roformer.yaml)<br> |
|
Model_128_4: [dereverb-echo_128_4_4_mel_band_roformer_sdr_dry_12.4235.ckpt](./dereverb-echo_128_4_4_mel_band_roformer_sdr_dry_12.4235.ckpt)<br> |
|
Instr dry sdr: 12.4235 |
|
|
|
Instruments: [dry, other]<br> |
|
Finetuned from: `model_mel_band_roformer_ep_3005_sdr_11.4360.ckpt`<br> |
|
Datasets: |
|
- Training datasets: 270 songs from [opencpop](https://github.com/wenet-e2e/opencpop) and [GTSinger](https://github.com/GTSinger/GTSinger) |
|
- Validation datasets: 30 songs from my own collection |
|
- All random reverbs and delay effects are generated by [this python script](./scripts/create_reverb_delay.py) and sorted into the mustb18 dataset format. |
|
|
|
### Thanks |
|
|
|
- Mel-Band-Roformer [[Paper](https://arxiv.org/abs/2310.01809), [Repository](https://github.com/lucidrains/BS-RoFormer)] |
|
- [ZFTurbo](https://github.com/ZFTurbo)'s training code [[Music-Source-Separation-Training](https://github.com/ZFTurbo/Music-Source-Separation-Training)] |
|
- [CN17161](https://github.com/CN17161) provided GPUs. |
|
- [Glucy-2](https://github.com/Glucy-2) provided technical assistance. |