I-Vector / X-Vector
Definition
Fixed-length mathematical representations of a speech utterance used in automatic speaker recognition. I-vectors are derived from Gaussian mixture model statistics; x-vectors are embeddings learned by a deep neural network. Both are scored by a PLDA back-end.
- I-vector basis
- Gaussian mixture model statistics
- X-vector basis
- Deep neural network embedding
- Common scoring back-end
- PLDA (probabilistic linear discriminant analysis)
- Field
- Automatic speaker recognition
Common questions
Does an automatic i-vector or x-vector score replace an expert's auditory-acoustic comparison?+
No, standards such as the ENFSI Best Practice Manual treat automatic scores as one input supporting the expert's overall opinion, not a standalone determination, because factors like channel mismatch and limited reference data can affect the automatic score's reliability.
Why do both i-vector and x-vector systems use a PLDA back-end?+
PLDA models the separate contributions of between-speaker and within-speaker variability, letting the system convert a raw vector similarity into a calibrated likelihood ratio that expresses the strength of the speaker comparison evidence.
Related terms
- Forensic Speaker Comparison
- A systematic examination comparing acoustic and phonetic features of a questioned voice recording against known reference recordings of a named individual, expressed...
- Formant Frequencies
- Resonance frequencies of the vocal tract that shape vowel quality. F1 and F2 (the first and second formants) are the most informative...
- IAFPA
- International Association for Forensic Phonetics and Acoustics: the primary professional body for forensic phoneticians and audio analysts, which publishes guidelines on speaker...
- Likelihood Ratio (LR)
- The ratio of two conditional probabilities: the probability of the observed evidence given the prosecution's hypothesis (same source), divided by the probability...
- PLDA (Probabilistic Linear Discriminant Analysis)
- A statistical back-end model used with i-vector and x-vector systems to compute a similarity score between two utterance representations, normalised for within-speaker...