Oxford Wave Research VOCALISE
VOCALISE is a forensic automatic speaker recognition desktop software for Windows 8 and above.
Voice Comparison and Analysis of the Likelihood of Speech Evidence Powerful and flexible voice biometrics software for performing fast and accurate speaker comparisons using Deep Neural Networks.
VOCALISE is used in forensic laboratories in more than 23 countries across the world.
Key features:
- It offers spectral and phonetic feature analysis
- Comparison of multi-speaker files without diarisation
- Select regions of speech in a multi-speaker file for comparison
- Import and convert files in a wide range of audio formats
- Command-line and API versions available
VOCALISE allows the user to compare multi-speaker files without the need for diarisation using segmental approach, or by easy selection of speech from speakers of interest in multi-speaker files. It operates using the state-of-the-art x-vector PLDA (Probabilistic Linear Discriminant Analysis) framework and offers powerful visualisation features for i-vector and x-vector approaches in the form of an interactive 3D point cloud of speakers. Score calibration enables interpretation of scores across comparisons in different conditions with condition adaptation for ‘tuning’ the i-vector or x-vector PLDA system to a particular or unseen recording condition.
VOCALISE provides three measures of objective audio quality and an overall ‘star rating’ for each file and allows the user to sort files based on objective quality metrics and send subsets of files for analysis/comparison or write them to a new folder.
VOCALISE is supplied with Bio-Metrics software that provides performance assessments via a multitude of approaches, such as equal error rates and DET curves, zooplots, and calculation of likelihood ratios under same-speaker and different-speaker hypotheses