Skip to content

GitLab

"vscode:/vscode.git/clone" did not exist on "1f9078d6aef1ce5de4efdfcc12b56405d8c1fa0d"
Unverified Commit b4284de0 authored by Bhargav Kathivarapu's avatar Bhargav Kathivarapu Committed by GitHub
Browse files

Fix Kaldi mfcc device/dtype and migrate test (#681) 



* Fix device/dtype compatibility of Kaldi mfcc

* Migrate Kaldi mfcc test

* Remove failing tests
Signed-off-by: default avatarBhargav Kathivarapu <bhargavkathivarapu31@gmail.com>
parent a466b3c2
Show whitespace changes
Inline Side-by-side
Showing with 132 additions and 2 deletions
test/assets/kaldi_test_mfcc_args.json 0 → 100644
View file @ b4284de0
{"blackman_coeff": 0.0063, "energy_floor": 3.0323, "frame_length": 0.625, "frame_shift": 0.0625, "high_freq": 5842, "htk_compat": false, "low_freq": 3240, "num_mel_bins": 4, "preemphasis_coefficient": 0.29, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 73.5792, "vtln_high": 5749, "vtln_low": 4449, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.013, "energy_floor": 1.8509, "frame_length": 1.1875, "frame_shift": 0.625, "high_freq": 7999, "htk_compat": false, "low_freq": 4330, "num_mel_bins": 5, "preemphasis_coefficient": 0.38, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 8.1048, "vtln_high": 7497, "vtln_low": 7397, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.0487, "energy_floor": 1.3641, "frame_length": 1.0, "frame_shift": 0.8125, "high_freq": 7892, "htk_compat": true, "low_freq": 1904, "num_mel_bins": 8, "preemphasis_coefficient": 0.26, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 5, "cepstral_lifter": 34.0918, "vtln_high": 4400, "vtln_low": 2737, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 0.0577, "energy_floor": 2.4313, "frame_length": 1.0625, "frame_shift": 0.875, "high_freq": 2922, "htk_compat": true, "low_freq": 274, "num_mel_bins": 6, "preemphasis_coefficient": 0.48, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 5, "cepstral_lifter": 21.3007, "vtln_high": 1352, "vtln_low": 280, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 0.0718, "energy_floor": 1.3071, "frame_length": 1.1875, "frame_shift": 0.5, "high_freq": 3159, "htk_compat": true, "low_freq": 759, "num_mel_bins": 8, "preemphasis_coefficient": 0.04, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 6, "cepstral_lifter": 2.6493, "vtln_high": 3145, "vtln_low": 3119, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 0.083, "energy_floor": 2.1607, "frame_length": 0.75, "frame_shift": 0.75, "high_freq": 5872, "htk_compat": true, "low_freq": 708, "num_mel_bins": 5, "preemphasis_coefficient": 0.95, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 24.5097, "vtln_high": 5231, "vtln_low": 3888, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 0.0933, "energy_floor": 1.577, "frame_length": 1.1875, "frame_shift": 1.1875, "high_freq": 7519, "htk_compat": false, "low_freq": 357, "num_mel_bins": 4, "preemphasis_coefficient": 0.5, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 88.0941, "vtln_high": 7042, "vtln_low": 5298, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 0.1048, "energy_floor": 0.5013, "frame_length": 0.75, "frame_shift": 1.0625, "high_freq": 6426, "htk_compat": true, "low_freq": 3613, "num_mel_bins": 5, "preemphasis_coefficient": 0.96, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 73.5838, "vtln_high": 5816, "vtln_low": 3997, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 0.1447, "energy_floor": 4.7142, "frame_length": 1.0625, "frame_shift": 0.4375, "high_freq": 7629, "htk_compat": true, "low_freq": 3498, "num_mel_bins": 7, "preemphasis_coefficient": 0.39, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 19.4145, "vtln_high": 7169, "vtln_low": 6751, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 0.1473, "energy_floor": 4.9154, "frame_length": 1.125, "frame_shift": 0.875, "high_freq": 3631, "htk_compat": false, "low_freq": 1229, "num_mel_bins": 8, "preemphasis_coefficient": 0.04, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 34.479, "vtln_high": 3390, "vtln_low": 1536, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.1485, "energy_floor": 4.275, "frame_length": 0.6875, "frame_shift": 0.75, "high_freq": 5222, "htk_compat": true, "low_freq": 311, "num_mel_bins": 5, "preemphasis_coefficient": 0.31, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 48.7991, "vtln_high": 4833, "vtln_low": 513, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 0.1829, "energy_floor": 4.5358, "frame_length": 0.9375, "frame_shift": 0.9375, "high_freq": 6148, "htk_compat": true, "low_freq": 455, "num_mel_bins": 5, "preemphasis_coefficient": 0.59, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 48.6724, "vtln_high": 3138, "vtln_low": 2247, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.2002, "energy_floor": 1.4805, "frame_length": 0.9375, "frame_shift": 0.875, "high_freq": 7621, "htk_compat": false, "low_freq": 2232, "num_mel_bins": 5, "preemphasis_coefficient": 0.03, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 5, "cepstral_lifter": 69.3653, "vtln_high": 7087, "vtln_low": 2800, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.2107, "energy_floor": 3.475, "frame_length": 1.125, "frame_shift": 0.8125, "high_freq": 5701, "htk_compat": true, "low_freq": 1629, "num_mel_bins": 4, "preemphasis_coefficient": 0.09, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 77.7066, "vtln_high": 5622, "vtln_low": 5544, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.2523, "energy_floor": 0.148, "frame_length": 1.0, "frame_shift": 1.125, "high_freq": 5833, "htk_compat": false, "low_freq": 556, "num_mel_bins": 4, "preemphasis_coefficient": 0.66, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 57.0398, "vtln_high": 4519, "vtln_low": 3600, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 0.2959, "energy_floor": 2.3729, "frame_length": 0.625, "frame_shift": 0.5, "high_freq": 6757, "htk_compat": false, "low_freq": 1744, "num_mel_bins": 6, "preemphasis_coefficient": 0.2, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 99.871, "vtln_high": 4957, "vtln_low": 3549, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 0.3642, "energy_floor": 3.5246, "frame_length": 0.4375, "frame_shift": 0.9375, "high_freq": 7942, "htk_compat": false, "low_freq": 3282, "num_mel_bins": 4, "preemphasis_coefficient": 0.52, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 74.4735, "vtln_high": 5601, "vtln_low": 4966, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.4891, "energy_floor": 1.989, "frame_length": 1.125, "frame_shift": 0.875, "high_freq": 3219, "htk_compat": true, "low_freq": 973, "num_mel_bins": 5, "preemphasis_coefficient": 0.91, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 85.357, "vtln_high": 3181, "vtln_low": 3129, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 0.5428, "energy_floor": 1.2368, "frame_length": 0.5625, "frame_shift": 1.0, "high_freq": 6700, "htk_compat": false, "low_freq": 749, "num_mel_bins": 4, "preemphasis_coefficient": 0.86, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 16.2782, "vtln_high": 5573, "vtln_low": 4988, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 0.5495, "energy_floor": 2.9502, "frame_length": 1.1875, "frame_shift": 1.1875, "high_freq": 3873, "htk_compat": true, "low_freq": 1564, "num_mel_bins": 5, "preemphasis_coefficient": 0.05, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 5, "cepstral_lifter": 59.0075, "vtln_high": 3870, "vtln_low": 3750, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 0.6457, "energy_floor": 2.0199, "frame_length": 0.875, "frame_shift": 0.8125, "high_freq": 6510, "htk_compat": false, "low_freq": 1482, "num_mel_bins": 4, "preemphasis_coefficient": 0.26, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 49.7663, "vtln_high": 5461, "vtln_low": 4039, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.7031, "energy_floor": 4.038, "frame_length": 1.125, "frame_shift": 0.6875, "high_freq": 6433, "htk_compat": true, "low_freq": 2336, "num_mel_bins": 8, "preemphasis_coefficient": 0.7, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 18.0061, "vtln_high": 5902, "vtln_low": 3191, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 0.7197, "energy_floor": 3.2075, "frame_length": 1.0625, "frame_shift": 0.25, "high_freq": 4448, "htk_compat": true, "low_freq": 378, "num_mel_bins": 4, "preemphasis_coefficient": 0.31, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 71.591, "vtln_high": 3497, "vtln_low": 3331, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 0.7238, "energy_floor": 1.9087, "frame_length": 1.1875, "frame_shift": 0.75, "high_freq": 5457, "htk_compat": true, "low_freq": 1775, "num_mel_bins": 7, "preemphasis_coefficient": 0.48, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 29.2858, "vtln_high": 5349, "vtln_low": 3987, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 0.7507, "energy_floor": 0.2754, "frame_length": 0.875, "frame_shift": 0.8125, "high_freq": 6405, "htk_compat": false, "low_freq": 1972, "num_mel_bins": 5, "preemphasis_coefficient": 0.83, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 52.5962, "vtln_high": 4597, "vtln_low": 4417, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 0.7954, "energy_floor": 0.3451, "frame_length": 1.1875, "frame_shift": 0.625, "high_freq": 4078, "htk_compat": false, "low_freq": 796, "num_mel_bins": 8, "preemphasis_coefficient": 0.47, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 8, "cepstral_lifter": 42.3128, "vtln_high": 2299, "vtln_low": 1094, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 0.8432, "energy_floor": 1.3765, "frame_length": 0.9375, "frame_shift": 1.1875, "high_freq": 6004, "htk_compat": true, "low_freq": 2302, "num_mel_bins": 4, "preemphasis_coefficient": 0.6, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 74.1116, "vtln_high": 4129, "vtln_low": 2898, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 0.8549, "energy_floor": 4.8924, "frame_length": 0.3125, "frame_shift": 1.125, "high_freq": 5643, "htk_compat": true, "low_freq": 956, "num_mel_bins": 4, "preemphasis_coefficient": 0.32, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 76.0384, "vtln_high": 2672, "vtln_low": 1762, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 0.9502, "energy_floor": 1.9738, "frame_length": 0.75, "frame_shift": 0.25, "high_freq": 7773, "htk_compat": true, "low_freq": 1205, "num_mel_bins": 7, "preemphasis_coefficient": 0.5, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 62.9038, "vtln_high": 7460, "vtln_low": 7174, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 1.0759, "energy_floor": 1.9132, "frame_length": 1.1875, "frame_shift": 0.625, "high_freq": 3529, "htk_compat": false, "low_freq": 227, "num_mel_bins": 8, "preemphasis_coefficient": 0.26, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 4.1559, "vtln_high": 1976, "vtln_low": 972, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 1.147, "energy_floor": 4.3972, "frame_length": 0.9375, "frame_shift": 0.75, "high_freq": 6393, "htk_compat": true, "low_freq": 2451, "num_mel_bins": 4, "preemphasis_coefficient": 0.06, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 67.4571, "vtln_high": 5460, "vtln_low": 3654, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 1.2069, "energy_floor": 0.4607, "frame_length": 1.125, "frame_shift": 0.5625, "high_freq": 5864, "htk_compat": true, "low_freq": 1512, "num_mel_bins": 7, "preemphasis_coefficient": 0.17, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 33.0194, "vtln_high": 5438, "vtln_low": 3920, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 1.3369, "energy_floor": 4.2619, "frame_length": 0.6875, "frame_shift": 0.375, "high_freq": 5307, "htk_compat": true, "low_freq": 666, "num_mel_bins": 6, "preemphasis_coefficient": 0.19, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 88.3186, "vtln_high": 4677, "vtln_low": 2590, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 1.3699, "energy_floor": 3.0236, "frame_length": 1.0625, "frame_shift": 0.75, "high_freq": 3720, "htk_compat": true, "low_freq": 1980, "num_mel_bins": 4, "preemphasis_coefficient": 0.13, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 57.6793, "vtln_high": 3441, "vtln_low": 3396, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 1.3763, "energy_floor": 1.6574, "frame_length": 1.125, "frame_shift": 0.8125, "high_freq": 2816, "htk_compat": false, "low_freq": 1021, "num_mel_bins": 4, "preemphasis_coefficient": 0.91, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 45.2819, "vtln_high": 2547, "vtln_low": 1123, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 1.4097, "energy_floor": 4.1523, "frame_length": 0.875, "frame_shift": 0.375, "high_freq": 6164, "htk_compat": false, "low_freq": 987, "num_mel_bins": 4, "preemphasis_coefficient": 0.06, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 75.0589, "vtln_high": 5873, "vtln_low": 5807, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 1.4295, "energy_floor": 3.7938, "frame_length": 1.125, "frame_shift": 0.75, "high_freq": 3382, "htk_compat": false, "low_freq": 471, "num_mel_bins": 4, "preemphasis_coefficient": 0.42, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 78.588, "vtln_high": 3299, "vtln_low": 2540, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 1.4677, "energy_floor": 4.0728, "frame_length": 1.125, "frame_shift": 1.0625, "high_freq": 7698, "htk_compat": true, "low_freq": 569, "num_mel_bins": 6, "preemphasis_coefficient": 0.5, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 76.8484, "vtln_high": 7453, "vtln_low": 7251, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 1.4979, "energy_floor": 1.1705, "frame_length": 1.1875, "frame_shift": 0.375, "high_freq": 4474, "htk_compat": true, "low_freq": 1123, "num_mel_bins": 7, "preemphasis_coefficient": 0.09, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 38.6407, "vtln_high": 3043, "vtln_low": 2934, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 1.555, "energy_floor": 1.8728, "frame_length": 0.875, "frame_shift": 0.9375, "high_freq": 5191, "htk_compat": true, "low_freq": 2262, "num_mel_bins": 4, "preemphasis_coefficient": 0.24, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 11.982, "vtln_high": 4607, "vtln_low": 4483, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 1.5626, "energy_floor": 3.7117, "frame_length": 1.125, "frame_shift": 0.125, "high_freq": 3008, "htk_compat": true, "low_freq": 534, "num_mel_bins": 5, "preemphasis_coefficient": 0.65, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 75.6661, "vtln_high": 2592, "vtln_low": 1621, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 1.5707, "energy_floor": 3.0409, "frame_length": 0.75, "frame_shift": 0.625, "high_freq": 7441, "htk_compat": true, "low_freq": 1554, "num_mel_bins": 6, "preemphasis_coefficient": 0.95, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 80.1563, "vtln_high": 7152, "vtln_low": 6151, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 1.5966, "energy_floor": 2.3442, "frame_length": 0.5625, "frame_shift": 0.5, "high_freq": 7944, "htk_compat": true, "low_freq": 1616, "num_mel_bins": 5, "preemphasis_coefficient": 0.49, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 80.8779, "vtln_high": 5720, "vtln_low": 4080, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 1.6229, "energy_floor": 2.0519, "frame_length": 1.1875, "frame_shift": 0.3125, "high_freq": 4871, "htk_compat": true, "low_freq": 1567, "num_mel_bins": 4, "preemphasis_coefficient": 0.79, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 42.9569, "vtln_high": 3483, "vtln_low": 3287, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 1.736, "energy_floor": 0.4063, "frame_length": 0.6875, "frame_shift": 0.0625, "high_freq": 6475, "htk_compat": true, "low_freq": 4439, "num_mel_bins": 4, "preemphasis_coefficient": 0.23, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 30.0984, "vtln_high": 5450, "vtln_low": 4909, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 1.7411, "energy_floor": 2.0918, "frame_length": 1.0625, "frame_shift": 0.8125, "high_freq": 6107, "htk_compat": true, "low_freq": 2523, "num_mel_bins": 4, "preemphasis_coefficient": 0.69, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 92.6839, "vtln_high": 5085, "vtln_low": 4771, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 1.7439, "energy_floor": 2.3782, "frame_length": 0.875, "frame_shift": 1.1875, "high_freq": 7669, "htk_compat": false, "low_freq": 4499, "num_mel_bins": 4, "preemphasis_coefficient": 0.81, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 95.7035, "vtln_high": 7521, "vtln_low": 7417, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 1.7611, "energy_floor": 4.2965, "frame_length": 0.8125, "frame_shift": 0.6875, "high_freq": 6607, "htk_compat": false, "low_freq": 454, "num_mel_bins": 7, "preemphasis_coefficient": 0.35, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 17.8265, "vtln_high": 6387, "vtln_low": 6105, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 1.7893, "energy_floor": 1.8005, "frame_length": 0.625, "frame_shift": 0.375, "high_freq": 2791, "htk_compat": true, "low_freq": 617, "num_mel_bins": 4, "preemphasis_coefficient": 0.96, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 93.405, "vtln_high": 1751, "vtln_low": 1690, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 1.8392, "energy_floor": 4.3711, "frame_length": 0.9375, "frame_shift": 0.75, "high_freq": 6978, "htk_compat": true, "low_freq": 453, "num_mel_bins": 4, "preemphasis_coefficient": 0.48, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 42.9768, "vtln_high": 6315, "vtln_low": 3995, "vtln_warp": 1.1059, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 1.9561, "energy_floor": 0.8419, "frame_length": 0.8125, "frame_shift": 1.125, "high_freq": 5308, "htk_compat": false, "low_freq": 1471, "num_mel_bins": 5, "preemphasis_coefficient": 0.62, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 51.0514, "vtln_high": 5221, "vtln_low": 5071, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 1.998, "energy_floor": 1.6949, "frame_length": 1.125, "frame_shift": 0.8125, "high_freq": 4678, "htk_compat": true, "low_freq": 2340, "num_mel_bins": 5, "preemphasis_coefficient": 0.44, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 70.1988, "vtln_high": 4041, "vtln_low": 2424, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 2.0106, "energy_floor": 4.5392, "frame_length": 0.6875, "frame_shift": 0.1875, "high_freq": 4776, "htk_compat": true, "low_freq": 1297, "num_mel_bins": 5, "preemphasis_coefficient": 0.14, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 37.5837, "vtln_high": 3995, "vtln_low": 2991, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 2.0835, "energy_floor": 2.8454, "frame_length": 0.9375, "frame_shift": 0.3125, "high_freq": 7496, "htk_compat": false, "low_freq": 1207, "num_mel_bins": 6, "preemphasis_coefficient": 0.76, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 6, "cepstral_lifter": 4.5734, "vtln_high": 3935, "vtln_low": 3932, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 2.2801, "energy_floor": 1.7051, "frame_length": 1.1875, "frame_shift": 0.125, "high_freq": 7958, "htk_compat": true, "low_freq": 561, "num_mel_bins": 4, "preemphasis_coefficient": 0.9, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 60.4649, "vtln_high": 7218, "vtln_low": 5709, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 2.315, "energy_floor": 0.4964, "frame_length": 0.5, "frame_shift": 0.4375, "high_freq": 6582, "htk_compat": false, "low_freq": 1010, "num_mel_bins": 4, "preemphasis_coefficient": 0.98, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 71.516, "vtln_high": 6157, "vtln_low": 4430, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 2.3236, "energy_floor": 0.7825, "frame_length": 0.8125, "frame_shift": 0.25, "high_freq": 7488, "htk_compat": true, "low_freq": 1363, "num_mel_bins": 4, "preemphasis_coefficient": 0.3, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 68.9678, "vtln_high": 3555, "vtln_low": 1851, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 2.3805, "energy_floor": 2.934, "frame_length": 0.75, "frame_shift": 0.25, "high_freq": 6076, "htk_compat": true, "low_freq": 80, "num_mel_bins": 4, "preemphasis_coefficient": 0.85, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 31.0805, "vtln_high": 2257, "vtln_low": 1533, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 2.4091, "energy_floor": 2.8812, "frame_length": 1.125, "frame_shift": 0.9375, "high_freq": 6086, "htk_compat": false, "low_freq": 1210, "num_mel_bins": 5, "preemphasis_coefficient": 0.59, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 96.1612, "vtln_high": 4840, "vtln_low": 1905, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 2.4134, "energy_floor": 2.6379, "frame_length": 1.1875, "frame_shift": 0.375, "high_freq": 3318, "htk_compat": false, "low_freq": 770, "num_mel_bins": 5, "preemphasis_coefficient": 0.6, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 73.9427, "vtln_high": 2044, "vtln_low": 1481, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 2.5228, "energy_floor": 3.1056, "frame_length": 1.125, "frame_shift": 1.1875, "high_freq": 5422, "htk_compat": false, "low_freq": 2825, "num_mel_bins": 7, "preemphasis_coefficient": 0.88, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 7, "cepstral_lifter": 4.6719, "vtln_high": 5337, "vtln_low": 5243, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 2.5577, "energy_floor": 0.7393, "frame_length": 0.8125, "frame_shift": 0.5, "high_freq": 5291, "htk_compat": true, "low_freq": 1445, "num_mel_bins": 5, "preemphasis_coefficient": 0.01, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 5.8944, "vtln_high": 4338, "vtln_low": 4330, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 2.5854, "energy_floor": 3.2219, "frame_length": 0.875, "frame_shift": 0.4375, "high_freq": 6924, "htk_compat": false, "low_freq": 4024, "num_mel_bins": 4, "preemphasis_coefficient": 1.0, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 0.0578, "vtln_high": 5707, "vtln_low": 5025, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 2.6674, "energy_floor": 2.777, "frame_length": 1.0625, "frame_shift": 0.3125, "high_freq": 3129, "htk_compat": true, "low_freq": 1706, "num_mel_bins": 4, "preemphasis_coefficient": 0.91, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 50.9241, "vtln_high": 2593, "vtln_low": 2198, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 2.6816, "energy_floor": 4.0548, "frame_length": 1.1875, "frame_shift": 0.625, "high_freq": 3182, "htk_compat": false, "low_freq": 157, "num_mel_bins": 6, "preemphasis_coefficient": 0.04, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 5, "cepstral_lifter": 31.3652, "vtln_high": 1203, "vtln_low": 1174, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 2.7879, "energy_floor": 3.3482, "frame_length": 0.6875, "frame_shift": 0.375, "high_freq": 4262, "htk_compat": true, "low_freq": 150, "num_mel_bins": 4, "preemphasis_coefficient": 0.68, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 92.794, "vtln_high": 3276, "vtln_low": 1685, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 2.8683, "energy_floor": 3.8162, "frame_length": 1.125, "frame_shift": 0.375, "high_freq": 6620, "htk_compat": false, "low_freq": 3389, "num_mel_bins": 7, "preemphasis_coefficient": 0.83, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 5, "cepstral_lifter": 82.2365, "vtln_high": 5365, "vtln_low": 4579, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 2.869, "energy_floor": 3.2618, "frame_length": 1.1875, "frame_shift": 0.9375, "high_freq": 5646, "htk_compat": true, "low_freq": 491, "num_mel_bins": 8, "preemphasis_coefficient": 0.89, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 59.9812, "vtln_high": 5397, "vtln_low": 2639, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 2.9211, "energy_floor": 4.144, "frame_length": 0.75, "frame_shift": 0.375, "high_freq": 7210, "htk_compat": true, "low_freq": 3666, "num_mel_bins": 4, "preemphasis_coefficient": 0.93, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 94.5907, "vtln_high": 6682, "vtln_low": 4979, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 2.9464, "energy_floor": 0.6798, "frame_length": 1.125, "frame_shift": 0.0625, "high_freq": 4445, "htk_compat": true, "low_freq": 323, "num_mel_bins": 6, "preemphasis_coefficient": 0.46, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 6, "cepstral_lifter": 7.8133, "vtln_high": 3755, "vtln_low": 1137, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 2.9633, "energy_floor": 1.9565, "frame_length": 0.875, "frame_shift": 0.0625, "high_freq": 6835, "htk_compat": false, "low_freq": 649, "num_mel_bins": 5, "preemphasis_coefficient": 0.77, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 80.8871, "vtln_high": 6691, "vtln_low": 6581, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 2.9643, "energy_floor": 3.8883, "frame_length": 0.75, "frame_shift": 0.25, "high_freq": 7373, "htk_compat": false, "low_freq": 1698, "num_mel_bins": 4, "preemphasis_coefficient": 0.92, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 63.3036, "vtln_high": 6650, "vtln_low": 5500, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 2.9697, "energy_floor": 2.0241, "frame_length": 1.125, "frame_shift": 0.6875, "high_freq": 2170, "htk_compat": false, "low_freq": 180, "num_mel_bins": 5, "preemphasis_coefficient": 0.28, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 95.8111, "vtln_high": 1266, "vtln_low": 521, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.0358, "energy_floor": 1.7295, "frame_length": 1.1875, "frame_shift": 1.0, "high_freq": 7222, "htk_compat": true, "low_freq": 858, "num_mel_bins": 4, "preemphasis_coefficient": 0.16, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 64.7537, "vtln_high": 6220, "vtln_low": 5229, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 3.0421, "energy_floor": 3.3343, "frame_length": 1.0, "frame_shift": 0.9375, "high_freq": 6477, "htk_compat": false, "low_freq": 1402, "num_mel_bins": 5, "preemphasis_coefficient": 0.99, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 26.1743, "vtln_high": 6381, "vtln_low": 5017, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 3.0919, "energy_floor": 4.5103, "frame_length": 0.625, "frame_shift": 1.0, "high_freq": 5323, "htk_compat": true, "low_freq": 937, "num_mel_bins": 5, "preemphasis_coefficient": 0.95, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 5, "cepstral_lifter": 82.2405, "vtln_high": 5130, "vtln_low": 5086, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 3.1463, "energy_floor": 4.5068, "frame_length": 0.6875, "frame_shift": 0.3125, "high_freq": 7587, "htk_compat": true, "low_freq": 3542, "num_mel_bins": 7, "preemphasis_coefficient": 0.78, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 28.5808, "vtln_high": 7478, "vtln_low": 7326, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.2007, "energy_floor": 1.3677, "frame_length": 1.125, "frame_shift": 0.1875, "high_freq": 4726, "htk_compat": true, "low_freq": 2027, "num_mel_bins": 4, "preemphasis_coefficient": 0.74, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 3.5732, "vtln_high": 4583, "vtln_low": 3742, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 3.2416, "energy_floor": 1.0604, "frame_length": 0.875, "frame_shift": 0.5, "high_freq": 4730, "htk_compat": false, "low_freq": 968, "num_mel_bins": 4, "preemphasis_coefficient": 0.11, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 60.5751, "vtln_high": 3542, "vtln_low": 1943, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.2698, "energy_floor": 3.0361, "frame_length": 1.0625, "frame_shift": 0.75, "high_freq": 4870, "htk_compat": true, "low_freq": 1281, "num_mel_bins": 7, "preemphasis_coefficient": 0.64, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 28.536, "vtln_high": 4401, "vtln_low": 3315, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.3078, "energy_floor": 4.9217, "frame_length": 1.0, "frame_shift": 0.3125, "high_freq": 6758, "htk_compat": true, "low_freq": 760, "num_mel_bins": 5, "preemphasis_coefficient": 0.98, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 97.4694, "vtln_high": 6022, "vtln_low": 5650, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 3.3206, "energy_floor": 0.023, "frame_length": 1.0625, "frame_shift": 0.5625, "high_freq": 5744, "htk_compat": true, "low_freq": 3901, "num_mel_bins": 5, "preemphasis_coefficient": 0.94, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 47.6031, "vtln_high": 5741, "vtln_low": 5524, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 3.4022, "energy_floor": 1.2172, "frame_length": 0.875, "frame_shift": 0.375, "high_freq": 7737, "htk_compat": false, "low_freq": 612, "num_mel_bins": 5, "preemphasis_coefficient": 0.35, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 65.1166, "vtln_high": 6852, "vtln_low": 5820, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.4339, "energy_floor": 2.6197, "frame_length": 1.125, "frame_shift": 0.1875, "high_freq": 3341, "htk_compat": true, "low_freq": 1275, "num_mel_bins": 7, "preemphasis_coefficient": 0.41, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 76.6062, "vtln_high": 3005, "vtln_low": 1680, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 3.4369, "energy_floor": 3.9198, "frame_length": 0.9375, "frame_shift": 0.125, "high_freq": 6218, "htk_compat": true, "low_freq": 904, "num_mel_bins": 5, "preemphasis_coefficient": 0.47, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 5, "cepstral_lifter": 92.8036, "vtln_high": 4870, "vtln_low": 1901, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.4557, "energy_floor": 1.5553, "frame_length": 0.75, "frame_shift": 0.375, "high_freq": 6642, "htk_compat": true, "low_freq": 1530, "num_mel_bins": 4, "preemphasis_coefficient": 0.72, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 28.8828, "vtln_high": 4490, "vtln_low": 2980, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 3.4753, "energy_floor": 4.7166, "frame_length": 0.75, "frame_shift": 0.3125, "high_freq": 7637, "htk_compat": true, "low_freq": 4992, "num_mel_bins": 4, "preemphasis_coefficient": 0.92, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 2.5456, "vtln_high": 6925, "vtln_low": 5486, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 3.5134, "energy_floor": 2.0285, "frame_length": 0.625, "frame_shift": 0.1875, "high_freq": 5229, "htk_compat": false, "low_freq": 595, "num_mel_bins": 4, "preemphasis_coefficient": 0.65, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 30.6069, "vtln_high": 5090, "vtln_low": 3467, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 3.5212, "energy_floor": 3.8251, "frame_length": 0.875, "frame_shift": 0.3125, "high_freq": 4092, "htk_compat": true, "low_freq": 545, "num_mel_bins": 5, "preemphasis_coefficient": 0.09, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 2.9422, "vtln_high": 1634, "vtln_low": 1000, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 3.5261, "energy_floor": 2.0251, "frame_length": 0.875, "frame_shift": 0.625, "high_freq": 7926, "htk_compat": false, "low_freq": 3916, "num_mel_bins": 7, "preemphasis_coefficient": 0.45, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 48.8818, "vtln_high": 7889, "vtln_low": 7527, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 3.539, "energy_floor": 1.6456, "frame_length": 1.125, "frame_shift": 0.1875, "high_freq": 5425, "htk_compat": true, "low_freq": 2326, "num_mel_bins": 7, "preemphasis_coefficient": 0.72, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 39.4555, "vtln_high": 4290, "vtln_low": 2715, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 3.5643, "energy_floor": 2.2424, "frame_length": 1.1875, "frame_shift": 0.875, "high_freq": 2140, "htk_compat": true, "low_freq": 59, "num_mel_bins": 4, "preemphasis_coefficient": 0.98, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 36.7118, "vtln_high": 1463, "vtln_low": 1358, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 3.5959, "energy_floor": 4.8866, "frame_length": 1.125, "frame_shift": 1.0625, "high_freq": 5150, "htk_compat": false, "low_freq": 3697, "num_mel_bins": 4, "preemphasis_coefficient": 0.46, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 74.0966, "vtln_high": 4277, "vtln_low": 3777, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 3.7223, "energy_floor": 3.4282, "frame_length": 1.0, "frame_shift": 0.125, "high_freq": 6601, "htk_compat": true, "low_freq": 1923, "num_mel_bins": 6, "preemphasis_coefficient": 0.05, "raw_energy": false, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 19.2839, "vtln_high": 6596, "vtln_low": 6594, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 3.7376, "energy_floor": 0.2093, "frame_length": 1.1875, "frame_shift": 0.75, "high_freq": 7830, "htk_compat": true, "low_freq": 4448, "num_mel_bins": 5, "preemphasis_coefficient": 0.27, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 5, "cepstral_lifter": 5.5865, "vtln_high": 5459, "vtln_low": 5056, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.812, "energy_floor": 0.4393, "frame_length": 0.75, "frame_shift": 1.0625, "high_freq": 5917, "htk_compat": false, "low_freq": 1272, "num_mel_bins": 4, "preemphasis_coefficient": 0.97, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 91.4723, "vtln_high": 3532, "vtln_low": 3056, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 3.8613, "energy_floor": 4.6574, "frame_length": 1.125, "frame_shift": 1.0, "high_freq": 3399, "htk_compat": true, "low_freq": 1576, "num_mel_bins": 5, "preemphasis_coefficient": 0.71, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 3, "cepstral_lifter": 29.1497, "vtln_high": 2440, "vtln_low": 1852, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 3.9117, "energy_floor": 4.6803, "frame_length": 0.5625, "frame_shift": 0.625, "high_freq": 5009, "htk_compat": false, "low_freq": 2542, "num_mel_bins": 4, "preemphasis_coefficient": 0.25, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 65.8362, "vtln_high": 4734, "vtln_low": 3050, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 4.1851, "energy_floor": 3.5211, "frame_length": 1.125, "frame_shift": 0.875, "high_freq": 4768, "htk_compat": false, "low_freq": 562, "num_mel_bins": 4, "preemphasis_coefficient": 0.05, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 36.961, "vtln_high": 1982, "vtln_low": 741, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 4.2197, "energy_floor": 3.7252, "frame_length": 0.9375, "frame_shift": 0.8125, "high_freq": 7453, "htk_compat": true, "low_freq": 1561, "num_mel_bins": 4, "preemphasis_coefficient": 0.06, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 44.78, "vtln_high": 6612, "vtln_low": 4074, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 4.2736, "energy_floor": 4.9552, "frame_length": 0.75, "frame_shift": 1.0, "high_freq": 5145, "htk_compat": false, "low_freq": 1705, "num_mel_bins": 4, "preemphasis_coefficient": 0.33, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 70.9332, "vtln_high": 4857, "vtln_low": 2223, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 4.3762, "energy_floor": 4.7209, "frame_length": 0.9375, "frame_shift": 0.0625, "high_freq": 5564, "htk_compat": true, "low_freq": 712, "num_mel_bins": 4, "preemphasis_coefficient": 0.74, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 39.2887, "vtln_high": 4353, "vtln_low": 3521, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 4.4229, "energy_floor": 0.4222, "frame_length": 1.0625, "frame_shift": 1.1875, "high_freq": 7822, "htk_compat": true, "low_freq": 4837, "num_mel_bins": 5, "preemphasis_coefficient": 0.04, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 5, "cepstral_lifter": 36.9181, "vtln_high": 7261, "vtln_low": 5703, "vtln_warp": 1.0, "window_type": "blackman", "dither": 0.0}
{"blackman_coeff": 4.4663, "energy_floor": 3.5767, "frame_length": 1.125, "frame_shift": 1.125, "high_freq": 5844, "htk_compat": false, "low_freq": 799, "num_mel_bins": 7, "preemphasis_coefficient": 0.37, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 34.2098, "vtln_high": 4554, "vtln_low": 1148, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 4.53, "energy_floor": 3.1492, "frame_length": 1.0625, "frame_shift": 0.375, "high_freq": 7706, "htk_compat": false, "low_freq": 3813, "num_mel_bins": 6, "preemphasis_coefficient": 0.74, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 2, "cepstral_lifter": 71.8337, "vtln_high": 7672, "vtln_low": 5265, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 4.5474, "energy_floor": 0.7883, "frame_length": 0.5625, "frame_shift": 1.0, "high_freq": 7283, "htk_compat": false, "low_freq": 2418, "num_mel_bins": 4, "preemphasis_coefficient": 0.68, "raw_energy": true, "remove_dc_offset": true, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 4, "cepstral_lifter": 70.0635, "vtln_high": 7277, "vtln_low": 7265, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 4.5663, "energy_floor": 1.127, "frame_length": 1.125, "frame_shift": 0.8125, "high_freq": 6069, "htk_compat": true, "low_freq": 167, "num_mel_bins": 8, "preemphasis_coefficient": 0.68, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": false, "use_energy": true, "num_ceps": 6, "cepstral_lifter": 1.624, "vtln_high": 2148, "vtln_low": 461, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 4.5896, "energy_floor": 2.7617, "frame_length": 1.0625, "frame_shift": 0.8125, "high_freq": 3851, "htk_compat": true, "low_freq": 1115, "num_mel_bins": 4, "preemphasis_coefficient": 0.03, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 49.7637, "vtln_high": 2897, "vtln_low": 2701, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 4.6128, "energy_floor": 0.1203, "frame_length": 1.1875, "frame_shift": 0.9375, "high_freq": 6901, "htk_compat": false, "low_freq": 3577, "num_mel_bins": 6, "preemphasis_coefficient": 0.25, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 70.5509, "vtln_high": 5962, "vtln_low": 4190, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 4.6262, "energy_floor": 4.1656, "frame_length": 1.1875, "frame_shift": 0.8125, "high_freq": 6147, "htk_compat": false, "low_freq": 1684, "num_mel_bins": 6, "preemphasis_coefficient": 0.58, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": true, "use_energy": true, "num_ceps": 5, "cepstral_lifter": 54.2056, "vtln_high": 5259, "vtln_low": 2363, "vtln_warp": 1.0, "window_type": "rectangular", "dither": 0.0}
{"blackman_coeff": 4.6741, "energy_floor": 4.3867, "frame_length": 1.125, "frame_shift": 1.125, "high_freq": 6273, "htk_compat": false, "low_freq": 2481, "num_mel_bins": 4, "preemphasis_coefficient": 0.15, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 75.6122, "vtln_high": 3701, "vtln_low": 2992, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 4.6765, "energy_floor": 1.2644, "frame_length": 1.125, "frame_shift": 0.75, "high_freq": 5204, "htk_compat": false, "low_freq": 276, "num_mel_bins": 4, "preemphasis_coefficient": 0.04, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": true, "subtract_mean": true, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 96.116, "vtln_high": 5148, "vtln_low": 2541, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 4.7216, "energy_floor": 2.4818, "frame_length": 0.8125, "frame_shift": 0.375, "high_freq": 6723, "htk_compat": true, "low_freq": 2352, "num_mel_bins": 6, "preemphasis_coefficient": 0.14, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 3, "cepstral_lifter": 32.0303, "vtln_high": 5598, "vtln_low": 2579, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 4.7919, "energy_floor": 2.6435, "frame_length": 0.625, "frame_shift": 0.5, "high_freq": 7971, "htk_compat": false, "low_freq": 1812, "num_mel_bins": 4, "preemphasis_coefficient": 0.65, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": true, "use_energy": false, "num_ceps": 4, "cepstral_lifter": 27.7648, "vtln_high": 7735, "vtln_low": 7419, "vtln_warp": 1.0, "window_type": "povey", "dither": 0.0}
{"blackman_coeff": 4.814, "energy_floor": 0.468, "frame_length": 1.0625, "frame_shift": 0.6875, "high_freq": 5252, "htk_compat": true, "low_freq": 569, "num_mel_bins": 6, "preemphasis_coefficient": 0.85, "raw_energy": true, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": true, "subtract_mean": false, "use_energy": false, "num_ceps": 6, "cepstral_lifter": 56.449, "vtln_high": 4397, "vtln_low": 4332, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
{"blackman_coeff": 4.95, "energy_floor": 4.5916, "frame_length": 1.125, "frame_shift": 1.0625, "high_freq": 5044, "htk_compat": true, "low_freq": 617, "num_mel_bins": 8, "preemphasis_coefficient": 0.89, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": false, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 8, "cepstral_lifter": 23.3238, "vtln_high": 2732, "vtln_low": 2677, "vtln_warp": 1.0, "window_type": "hanning", "dither": 0.0}
{"blackman_coeff": 4.9663, "energy_floor": 4.7867, "frame_length": 1.1875, "frame_shift": 0.5, "high_freq": 2424, "htk_compat": false, "low_freq": 350, "num_mel_bins": 4, "preemphasis_coefficient": 0.39, "raw_energy": false, "remove_dc_offset": false, "round_to_power_of_two": true, "snip_edges": false, "subtract_mean": true, "use_energy": true, "num_ceps": 2, "cepstral_lifter": 59.4319, "vtln_high": 1202, "vtln_low": 1063, "vtln_warp": 1.0, "window_type": "hamming", "dither": 0.0}
test/kaldi_compatibility_impl.py
View file @ b4284de0
...@@ -85,3 +85,14 @@ class Kaldi(common_utils.TestBaseMixin): ...@@ -85,3 +85,14 @@ class Kaldi(common_utils.TestBaseMixin):
command = ['compute-fbank-feats'] + _convert_args(**kwargs) + ['scp:-', 'ark:-'] command = ['compute-fbank-feats'] + _convert_args(**kwargs) + ['scp:-', 'ark:-']
kaldi_result = _run_kaldi(command, 'scp', wave_file) kaldi_result = _run_kaldi(command, 'scp', wave_file)
self.assert_equal(result, expected=kaldi_result, rtol=1e-4, atol=1e-8) self.assert_equal(result, expected=kaldi_result, rtol=1e-4, atol=1e-8)
@parameterized.expand(_load_params(common_utils.get_asset_path('kaldi_test_mfcc_args.json')))
@unittest.skipIf(_not_available('compute-mfcc-feats'), '`compute-mfcc-feats` not available')
def test_mfcc(self, kwargs):
"""mfcc should be numerically compatible with compute-mfcc-feats"""
wave_file = common_utils.get_asset_path('kaldi_file.wav')
waveform = torchaudio.load_wav(wave_file)[0].to(dtype=self.dtype, device=self.device)
result = torchaudio.compliance.kaldi.mfcc(waveform, **kwargs)
command = ['compute-mfcc-feats'] + _convert_args(**kwargs) + ['scp:-', 'ark:-']
kaldi_result = _run_kaldi(command, 'scp', wave_file)
self.assert_equal(result, expected=kaldi_result, rtol=1e-4, atol=1e-8)
torchaudio/compliance/kaldi.py
View file @ b4284de0
...@@ -696,6 +696,8 @@ def mfcc( ...@@ -696,6 +696,8 @@ def mfcc(
""" """
assert num_ceps <= num_mel_bins, 'num_ceps cannot be larger than num_mel_bins: %d vs %d' % (num_ceps, num_mel_bins) assert num_ceps <= num_mel_bins, 'num_ceps cannot be larger than num_mel_bins: %d vs %d' % (num_ceps, num_mel_bins)
device, dtype = waveform.device, waveform.dtype
# The mel_energies should not be squared (use_power=True), not have mean subtracted # The mel_energies should not be squared (use_power=True), not have mean subtracted
# (subtract_mean=False), and use log (use_log_fbank=True). # (subtract_mean=False), and use log (use_log_fbank=True).
# size (m, num_mel_bins + use_energy) # size (m, num_mel_bins + use_energy)
...@@ -717,7 +719,7 @@ def mfcc( ...@@ -717,7 +719,7 @@ def mfcc(
feature = feature[:, mel_offset:(num_mel_bins + mel_offset)] feature = feature[:, mel_offset:(num_mel_bins + mel_offset)]
# size (num_mel_bins, num_ceps) # size (num_mel_bins, num_ceps)
dct_matrix = _get_dct_matrix(num_ceps, num_mel_bins) dct_matrix = _get_dct_matrix(num_ceps, num_mel_bins).to(dtype=dtype, device=device)
# size (m, num_ceps) # size (m, num_ceps)
feature = feature.matmul(dct_matrix) feature = feature.matmul(dct_matrix)
...@@ -725,7 +727,7 @@ def mfcc( ...@@ -725,7 +727,7 @@ def mfcc(
if cepstral_lifter != 0.0: if cepstral_lifter != 0.0:
# size (1, num_ceps) # size (1, num_ceps)
lifter_coeffs = _get_lifter_coeffs(num_ceps, cepstral_lifter).unsqueeze(0) lifter_coeffs = _get_lifter_coeffs(num_ceps, cepstral_lifter).unsqueeze(0)
feature *= lifter_coeffs feature *= lifter_coeffs.to(device=device, dtype=dtype)
# if use_energy then replace the last column for htk_compat == true else first column # if use_energy then replace the last column for htk_compat == true else first column
if use_energy: if use_energy:
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment