# UIS-RNN

## Overview

This is the library for the
*Unbounded Interleaved-State Recurrent Neural Network (UIS-RNN)* algorithm.
UIS-RNN solves the problem of segmenting and clustering sequential data
by learning from examples.

This algorithm was originally proposed in the paper Fully Supervised Speaker Diarization.

The work has been introduced by Google AI Blog.

## Disclaimer

This open source implementation is slightly different than the internal one which we used to produce the results in the paper, due to dependencies on some internal libraries.

We CANNOT share the data, code, or model for the speaker recognition system (d-vector embeddings) used in the paper, since the speaker recognition system heavily depends on Google's internal infrastructure and proprietary data.

**This library is NOT an official Google product.**

## Dependencies

This library depends on:

- python 3.5+
- numpy 1.15.1
- pytorch 0.4.0
- scipy 1.1.0 (for evaluation only)

## Getting Started

### Run the demo

To get started, simply run this command:

`python3 demo.py --train_iteration=1000 -l=0.001 -hl=100`

This will train a UIS-RNN model using `data/training_data.npz`

,
then store the model on disk, perform inference on `data/testing_data.npz`

,
print the inference results, and save the averaged accuracy in a text file.

PS. The files under `data/`

are manually generated *toy data*,
for demonstration purpose only.
These data are very simple, so we are supposed to get 100% accuracy on the
testing data.

### Run the tests

You can also verify the correctness of this library by running:

`sh run_tests.sh`

If you fork this library and make local changes, be sure to use these tests as a sanity check.

Besides, these tests are also great examples for learning
the APIs, especially `tests/integration_test.py`

.

## Core APIs

### Glossary

General Machine Learning | Speaker Diarization |
---|---|

Sequence | Utterance |

Observation | Embedding / d-vector |

Label / Cluster ID | Speaker |

### Model construction

All algorithms are implemented as the `UISRNN`

class. First, construct a
`UISRNN`

object by:

`model = UISRNN(args)`

The definitions of the args are described in `model/arguments.py`

.
See `model_parser`

.

### Training

Next, train the model by calling the `fit()`

function:

`model.fit(train_sequence, train_cluster_id, args)`

Here `train_sequence`

should be a 2-dim numpy array of type `float`

, for
the **concatenated** observation sequences. For speaker diarization, this
could be the
d-vector embeddings.

For example, if you have *M* training utterances,
and each utterance is a sequence of *L* embeddings. Each embedding is
a vector of *D* numbers. Then the shape of `train_sequence`

is *N * D*,
where *N = M * L*.

`train_cluster_id`

is a 1-dim list or numpy array of strings, of length *N*.
It is the **concatenated** ground truth labels of all training data. For
speaker diarization, these labels are the speaker identifiers for each
observation (*e.g.* d-vector).

Since we are concatenating observation sequences, it is important to note that,
ground truth labels in `train_cluster_id`

across different sequences are
supposed to be **globally unique**.

For example, if the set of labels in the first
sequence is `{'A', 'B', 'C'}`

, and the set of labels in the second sequence
is `{'B', 'C', 'D'}`

. Then before concatenation, we should rename them to
something like `{'1_A', '1_B', '1_C'}`

and `{'2_B', '2_C', '2_D'}`

,
unless `'B'`

and `'C'`

in the two sequences are meaningfully identical
(in speaker diarization, this means they are the same speakers across
utterances).

The reason we concatenate all training sequences is that, we will be resampling
and *block-wise* shuffling the training data as a **data augmentation**
process, such that we result in a robust model even when there is insufficient
number of training sequences.

The definitions of the args are described in `model/arguments.py`

.
See `training_parser`

.

### Prediction

Once we are done with the training, we can run the trained model to perform
inference on new sequences by calling the `predict()`

function:

`predicted_label = model.predict(test_sequence, args)`

Here `test_sequence`

should be a 2-dim numpy array of type `float`

,
corresponding to a **single** observation sequence.

The returned `predicted_label`

is a list of integers, with the same
length as `test_sequence`

.

The definitions of the args are described in `model/arguments.py`

.
See `inference_parser`

.

## Citations

Our paper is cited as:

```
@article{zhang2018fully,
title={Fully Supervised Speaker Diarization},
author={Zhang, Aonan and Wang, Quan and Zhu, Zhenyao and Paisley, John and Wang, Chong},
journal={arXiv preprint arXiv:1810.04719},
year={2018}
}
```

## References

### Baseline diarization system

To learn more about our baseline diarization system based on
*unsupervised clustering* algorithms, check out
this site.

Specifically, the ground truth labels for the NIST SRE 2000 dataset (Disk6 and Disk8) can be found here.

### Speaker recognizer/encoder

To learn more about our speaker embedding system, check out this site.

We are aware of several third-party implementations of this work:

Please use your own judgement to decide whether you want to use these implementations.

**We are NOT responsible for the correctness of any third-party implementations.**