det.sre: Compute Detection Error Tradeoff statistics from SRE opbject
In davidavdav/sretools.R: Compute error metrics and make DET plots for NIST-like Speaker Recognition Evaluations

Description Usage Arguments Details Value Author(s) References See Also Examples

This function computes the relevant statistics of a NIST-style Spreaker Recognition Evaluation trial set. The scores in the trial set are sorted, and Detection Error Tradeoff statistics are computed, including Cdet, minimum Cdet, equal error rate, Cllr and minimum Cllr. The resulting class can be plotted using plot.det().

1	det(x, cond)

`x`	A data frame of class `sre`
`cond`	An optional list of factors for which to compensate unballanced trial counts

The trial set x must be an SRE object, which is a data.frame of class “sre” containing at least the columns “score”, 'dec” (decision) and “target” (whether this trial was a target or non-target trial). Other columns in the data frame can be used to specify factors for conditioning, or other meta data. Typically, an SRE object is created by read.sre or read.tnt.

If cond is specified, the trials are split in subset according to the factors listed in cond. Then the weights of the subsets are equalized by weighting the trials in the subset inversely proportional to their frequency, see cond.table and the refererence therein.

The function returns a list containing the important DET statistics

`Cllr`	Cost of LLR
`Cllr.min`	Minimum Cllr, computed using (weighted) isotonic regression
`EER`	The equal Error Rate, computed using the Convex Hull method
`Cdet`	The detection error costs, given the decisions in the trials
`Cdet.min`	The minimum dectection costs.
`mt`	The mean value of target scores
`mn`	The mean value of non-target scores
`nt`	The number of target trials
`nn`	The number of non-target trials
`n`	The number of trials
`...`	... more DET statistics ...

David A. van Leeuwen

Alvin Martin et al, “The DET Curve in Assessment of Detection Task Performance,” Proc. Interspeech, 1895–1898 (1997). Niko Br\"ummer and Johan du Preez, “Application-independent evaluation of speaker detection,” Computer, Speech and Language 20, 230–275, (2006). David van Leeuwen and Niko Br\"ummer, “An Introduction to Application-Independent Evaluation of Speaker Recognition System,” LNCS 4343 (2007). Foster Provost and Tom Fawcett, “Analysis and Visualization of Classifier Performance: Comparison under Imprecise Class and Cost Distributions,” Third International Conference on Knowledge Discovery and Data Mining (1997).

read.sre, read.tnt, setDCF, plot.det

## Load example SRE data:
## RU submission to EVALITA speaker recognition applications track
data(ru.2009)
## inspect details of data frame
ru.2009[1,]
setDCF("evalita")
## look at TC6 train condition and TS2 test condition (easiest task:-)
x <- subset(ru.2009, mcond=="TC6" & tcond=="TS2")
## compute det statistics
d <- det(x)
summary(d)
## and plot results
plot(d, main="RU TC6 TS1 primary submission EVALITA 2009")