Skip to content
Multiobjective Optimization Training of PLDA for Speaker Verification
MATLAB Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
matlab
python
LICENSE
README.md

README.md

MOT-sGPLDA-SRE14

Multiobjective Optimization Training of PLDA for Speaker Verification

  1. prepare data, make directory ./data and ./temp
    put NIST SRE14 i-vector challenge offical data on "./data/", there are "development_data_labels.csv, dev_ivectors.csv, ivec14_sre_segment_key_release.tsv, ivec14_sre_trial_key_release.tsv, model_ivectors.csv, target_speaker_models.csv, test_ivectors.csv"

  2. run ./python/sre14_preprocess.py.
    It will generate "./temp/sre14.mat"

  3. run ./matlab/gplda_demo.m
    The script will read "./temp/sre14.mat", and the results are
    2.347, 2.456 (Development dataset, EER), 2.307 (Evaluation dataset, EER),
    0.264, 0.269 (Development dataset, MDCF), 0.261 (Evaluation dataset, MDCF).

  4. run ./matlab/moplda_demo.m
    The script will read "./temp/sre14.mat", and the results are
    2.040, 2.193 (Development dataset, EER), 1.931 (Evaluation dataset, EER),
    0.233, 0.239 (Development dataset, MDCF), 0.229 (Evaluation dataset, MDCF).

  5. some experiment results.

A. train lambda with development and train vectors

  1. factor experiment,
    A EER, D EER, E EER, A DCF, D DCF, E DCF, factor
  2. 2.531, 2.794, 2.354, 0.272, 0.277, 0.267, 1.1
  3. 2.554, 2.825, 2.336, 0.269, 0.272, 0.266, 1.2
  4. 2.456, 2.677, 2.176, 0.250, 0.255, 0.247, 1.3
  5. 2.331, 2.579, 2.199, 0.238, 0.241, 0.235, 1.4
  6. 2.207, 2.399, 2.099, 0.233, 0.235, 0.230, 1.5
  7. 2.082, 2.272, 1.940, 0.230, 0.235, 0.225, 1.6
  8. 2.040, 2.193, 1.931, 0.233, 0.239, 0.229, 1.7
  9. 2.057, 2.180, 1.973, 0.238, 0.244, 0.232, 1.8
  10. 2.136, 2.241, 2.049, 0.244, 0.250, 0.238, 1.9
  11. 2.108, 2.261, 1.995, 0.242, 0.249, 0.237, 2.0

B. train lambda with development vectors

  1. factor experiment,
    A EER, D EER, E EER, A DCF, D DCF, E DCF, factor
  2. 2.182, 2.426, 1.994, 0.241, 0.244, 0.237, 1.1
  3. 2.257, 2.438, 2.099, 0.237, 0.243, 0.232, 1.2
  4. 2.369, 2.487, 2.225, 0.240, 0.245, 0.236, 1.3
  5. 2.307, 2.426, 2.210, 0.238, 0.245, 0.233, 1.4
  6. 2.481, 2.610, 2.330, 0.264, 0.273, 0.257, 1.5
  7. 2.340, 2.425, 2.254, 0.253, 0.260, 0.246, 1.6
  8. 2.347, 2.487, 2.267, 0.253, 0.260, 0.247, 1.7
  9. 2.337, 2.395, 2.301, 0.264, 0.272, 0.257, 1.8
  10. 2.406, 2.456, 2.351, 0.266, 0.274, 0.260, 1.9
  11. 2.506, 2.549, 2.435, 0.275, 0.281, 0.269, 2.0

C. train lambda with development and train vectors

  1. dimension experiment,
    A EER, D EER, E EER, A DCF, D DCF, E DCF, dimension
  2. 3.522, 3.556, 3.505, 0.498, 0.507, 0.491, 50
  3. 2.306, 2.329, 2.267, 0.288, 0.294, 0.284, 100
  4. 2.032, 2.241, 1.919, 0.239, 0.245, 0.234, 150
  5. 2.040, 2.193, 1.931, 0.233, 0.239, 0.229, 200
  6. 2.070, 2.260, 1.918, 0.234, 0.240, 0.229, 250

PS:

  1. The sGPLDA demo was downloaded from https://github.com/wangwei2009/MSR-Identity-Toolkit-v1.0
  2. Anaconda3, Python3, require sklearn
  3. Matlab R2016a
  4. A EER: NIST SRE14 i-vector challenge all data EER
    D EER: NIST SRE14 i-vector challenge Development dataset EER
    E EER: NIST SRE14 i-vector challenge Evaluation dataset EER
  5. A DCF: NIST SRE14 i-vector challenge all data MinDCF14
    D DCF: NIST SRE14 i-vector challenge Development dataset MinDCF14
    E DCF: NIST SRE14 i-vector challenge Evaluation dataset MinDCF14
You can’t perform that action at this time.