README.md

---
license: apache-2.0
language:
- en
tags:
- hearing loss
- challenge
- signal processing
- source separation
- audio
- audio-to-audio
- NonCausal
---

# Cadenza Challenge: CAD2-Task2

A Causal separation model for the CAD2-Task2 system.

This model is an ensemble of the following instruments:
- Bassoon
- Cello
- Clarinet
- Flute
- Oboe
- Sax
- Viola
- Violin

Each model is based on the ConvTasNet (Kaituo XU) with multichannel support (Alexandre Defossez). 
* Parameters:
  * B: 256
  * C: 2
  * H: 512
  * L: 20
  * N: 256
  * P: 3
  * R: 3
  * X: 8
  * audio_channels: 2
  * causal: true
  * mask_nonlinear: relu
  * norm_type: cLN

## Dataset
The model was trained using EnsembleSet and CadenzaWoodwind datasets.

## How to use

```
from dynamic_source_separator import DynamicSourceSeparator

model = DynamicSourceSeparator.from_pretrained(
    "cadenzachallenge/Dynamic_Source_Separator_Causal"
).cpu()

```

## Description

Audio source separation model used in Sytem T002 for [Cadenza2 Task2 Challenge](https://cadenzachallenge.org/docs/cadenza2/Rebalancing%20Classical/rebalancing)

The model is a finetune of the 8 ConvTasNet models from the Task2 baseline. 
The training optimised the estimated sources and the recosntructed mixture

$$
  Loss = \sum_{}^{Sources}(L_1(estimated~source, ref~source)) + L_1(reconstructed~mixture, original~mixture)
$$
```Python
def dynamic_masked_loss(mixture, separated_sources, ground_truth_sources, indicator):
    # Reconstruction Loss
    reconstruction = sum(separated_sources.values())
    reconstruction_loss = nn.L1Loss()(reconstruction, mixture)
    # Separation Loss
    separation_loss = 0
    for instrument, active in indicator.items():
        if active:
            separation_loss += nn.L1Loss()(
                separated_sources[instrument], ground_truth_sources[instrument]
            )
    return reconstruction_loss + separation_loss
```
Model and T002 recipe are shared in [Clarity toolkit](https://github.com/claritychallenge/clarity)