Only for linear classifiers like XGB, SVM, SGR and SVR you have the possibility to disregard training and development splits and do a cross validation, i.e. validate one data set in a circular manner against itself.

The basic idea is that you take part of the data and evaluate against the rest, and in the next round take another part and so forth, until all data has been evaluated. Because the speaker identity is so strong in speech, this is done usually in a speaker exclusive manner, known under the term "leave one speaker out " (LOSO).

If you have too many speakers and/or each speaker really only one sample, you might want to split your speakers into groups and do a "leave one speaker group out" strategy (LOGO).

A related approach is known under the name k fold cross validation, where k usually equals 10.
When you only have one sample per speaker, this might make more sense.
So, how would you do that with Nkululeko?
First, you would define a training and development split for your data anyway, because Nkululeko is expecting it if there is only one database. You might set that to random, it's not used anyway:

[DATA]
mydata.split_strategy = random 

Then in your config file, you specify in the MODEL section either:

[MODEL] 
logo = 10 

to assign 10 groups to your speakers and then evaluate each group against all others.
If you want to do a leave-one-speaker_out experiment (LOSO), simply assign the number for logo the number of your speakers.

If there already is a fold column in your data, this will be used, otherwise Nkululeko will randomly assign folds to speakers.

Or you do

[MODEL] 
k_fold_cross = 5 

for instance to disregard speaker information and simply evaluate 5 times a fifth of the data against the rest.
We use stratified sets, i.e. the algorithm tries to balance the class data within each set.

Often you simply start an experiment with some audio data that you got from somewhere in no special format. Often the labels are encoded in the filenames.
If so, this Python script can help to convert the audio to a Nkululeko readable format and generate a CSV (comma separated values) file.

import os
from audeer import list_file_names
from os.path import basename

# folder with the original audio files (in wav format)
root = './orig_wav/'
# output folder, empty at the beginning
out_dir = './audio/'
# name of the output file list
out_file = 'data.csv'

# get a list of wav files
list = list_file_names(root, filetype = 'wav', basenames=True, recursive=True)
# write the list header (change to your data)
with open(out_file, 'a') as the_file:
    the_file.write('file,type\n')
# for each file
for file in list:
    # get the file name without path
    fn = basename(file)
    # convert to 16kHz sampling rate and mono channel 
    os.system(f'sox {root+file} -r 16000 -c 1 {out_dir+fn}')
    # extract the annotation label from the file name (change this to your needs)
    label = fn[0]
    # lastly: add file to list 
    with open(out_file, 'a') as the_file:
        the_file.write(f'{out_dir+fn},{label}\n')

The resulting data list can then be read by Nkululeko in the config file (using randomly 30 % of the data as development set):

[DATA]
my_data = /some_path/data.csv
my_data.type = csv
my_data.split_strategy = random
my_data.testsplit = 30

Since version 0.15.0, dropout within MLP (multi layer perceptron) models is supported by Nkululeko, here's an example:

[MODEL]
type = mlp
layers = {'l1':128, 'l2':16}
drop = .5
learning_rate = 0.0001

meaning, that after each hidden layer a dropout probability of 50 percent is applied.

Since version 0.13.0 with Nkululeko you can define all root folders for your databases at one single place.
This is very handy if you work in paralell on several computers, e.g. a development and a deployment environment.

In the [DATA] section of your ini file, you specify the path to the local data root folder file like this:

[DATA]
root_folders = data_roots.ini
databases = ['dataset_1']
...

and then within the data_roots.ini file (you can actually call it what you want), you declare the folders to your databases like this:

[DATA]
dataset_1 = /mypath/d1/
dataset_1.files_tables = ['files']
dataset_2 = ./d2
...

you can add all your data set options that you need in this file:

[DATA]
emodb = /mypath/d1/
emodb.split_strategy = speaker_split
emodb.testsplit = 40
emodb.mapping = {'anger':'angry', 'happiness':'happy', 'sadness':'sad', 'fear':'fright.', 'neutral':'neutral'}
dataset_2 = ./d2
dataset_2.files_tables = ['files_test', 'files_train']

If you define those fields in your experiment ini file, it will have precedence.

Nkululeko is a tool to ease machine learning on speech databases.
This tutorial should help you to import databases.
There are two formats upported:
1) csv (comma seperated values)
2) audformat

CSV format

The easiest is CSV, you simply create a table with the following informations:

• file: the path to the audio file
• task: is the speaker characteristics value that you want to explore, e.g. age or emotion, or both

and then fill it with values of your database. Optionally, your data can contain any amount of additional information in further columns. Some naming conventions are pre-defined:

• speaker: speaker id, a string being unique for samples from one speaker
• gender: biological sex
• age: an integer between 0 and 100 denoting the age in years.

So a file for emotion might look like this

file, speaker, gender, emotion
<path to>/s12343.wav, s1, female, happy
...

You can then specify the data in your initialization file like this:

[DATA]
databases = ['my_db']
my_db.type = csv
my_db = <path to>/my_data_file.csv
my_db.absolute_path = False 
...
target = emotion

You should set the flag absolute_path depending on whether

• the file paths start from the location of where you run Nkululeko (or start from root: /), then True
• or they start from the location where the data resides, then False

(if in doubt, just try it out: there should be an error message that the audio files don't exist)

You can not specify split tables with this format, but would have to simply split the file in several databases.

There is an example on how to import the ravdess database here.

And this would be an example ini file to use it:

[EXP]
root = ./tests/results/
name = exp_ravdess
runs = 1
epochs = 1
save = True
[DATA]
databases = ['train', 'test', 'dev']
train = ../nkululeko/data/ravdess/ravdess_train.csv
train.type = csv
train.absolute_path = False
train.split_strategy = train
dev = ../nkululeko/data/ravdess/ravdess_dev.csv
dev.type = csv
dev.absolute_path = False
dev.split_strategy = train
test = ../nkululeko/data/ravdess/ravdess_test.csv
test.type = csv
test.absolute_path = False
test.split_strategy = test
target = emotion
labels = ['angry', 'happy', 'neutral', 'sad']
[FEATS]
type = ['os']
scale = standard
[MODEL]
type = xgb

I.e. the splits train and dev get concatenated to a common train set

Fun fact: the result is:

audformat

audformat allows for many usecases, so the specification might be more complex.
So in the easiest case you have a database with two tables, one called files that contains the speaker informations (id and sex) and one called like your task (aka target), so for example age or emotion.
That's the case for our demo example, the Berlin EmoDB, ando so you can include it simply with.

[DATA]
databases = ['emodb']
emodb = /<path to>/emodb/
target = emotion
...

But if there are more tables and they have special names, you can specifiy them like this:

[DATA]
databases = ['msp']
# path to data
msp = /<path to>/msppodcast/
# tables with speaker information
msp.files_tables =  ['files.test-1', 'files.train']
# tables with task labels
msp.target_tables =  ['emotion.test-1', 'emotion.train']
# train and evaluation splits will be provided
msp.split_strategy = specified
# here are the test/evaluatoin split tables
msp.test_tables = ['emotion.test-1']
# here are the training tables
msp.train_tables = ['emotion.train']
target = emotion

Nkululeko supports classification and regression.
Classification means predicting a class (or category) from data, regression predicting a continuous value, as for example the speaker age in years.

If you want to use classification with continuous variables, you need to first bin it, which means that you put the values into pre-defined bins. To stay with our age example, you'd declare everyone above 50 years as old and all other as young.

This post shows you how to do that with Nkululeko by setting up your .ini file.

You set up the experiment as classification type:

[EXP]
...
type = classification

But declare the data to be continuous:

[DATA]
...
type = continuous
labels = ['u40', '40ies', '50ies', '60ies', 'ü70']
bins  = [-1000,  40, 50, 60, 70, 1000]

Then the data will be binned according to the sepecified bins and labeled accordingly.
You need (number of labels) + 1 values for the bins, as they are given lower and upper limit. It makes sense to set the lower and upper absolute limits extreme as you don't know what the classifier will predict.