Migrate kaldi fbank (#672)

* Migrate fbank tests

* Update CCI job environment

* Remove invalid test cases
Signed-off-by: Bhargav Kathivarapu <bhargavkathivarapu31@gmail.com>

.circleci/unittest/linux/scripts/environment.yml

@@ -13,3 +13,4 @@ dependencies:
     - clang-format
     - kaldi-io
     - scipy
+    - parameterized

.circleci/unittest/windows/scripts/environment.yml

@@ -14,3 +14,4 @@ dependencies:
     - PySoundFile
     - librosa
     - future
+    - parameterized

test/kaldi_compatibility_impl.py

 """Test suites for checking numerical compatibility against Kaldi"""
+import json
 import shutil
 import unittest
 import subprocess
@@ -9,6 +10,7 @@ import torchaudio.functional as F
 import torchaudio.compliance.kaldi
 
 from . import common_utils
+from parameterized import parameterized, param
 
 
 def _not_available(cmd):
@@ -47,6 +49,11 @@ def _run_kaldi(command, input_type, input_value):
     return torch.from_numpy(result.copy())  # copy supresses some torch warning
 
 
+def _load_params(path):
+    with open(path, 'r') as file:
+        return [param(json.loads(line)) for line in file]
+
+
 class Kaldi(common_utils.TestBaseMixin):
     def assert_equal(self, output, *, expected, rtol=None, atol=None):
         expected = expected.to(dtype=self.dtype, device=self.device)
@@ -68,34 +75,10 @@ class Kaldi(common_utils.TestBaseMixin):
         kaldi_result = _run_kaldi(command, 'ark', tensor)
         self.assert_equal(result, expected=kaldi_result)
 
+    @parameterized.expand(_load_params(common_utils.get_asset_path('kaldi_test_fbank_args.json')))
     @unittest.skipIf(_not_available('compute-fbank-feats'), '`compute-fbank-feats` not available')
-    def test_fbank(self):
+    def test_fbank(self, kwargs):
         """fbank should be numerically compatible with compute-fbank-feats"""
-        kwargs = {
-            'blackman_coeff': 4.3926,
-            'dither': 0.0,
-            'energy_floor': 2.0617,
-            'frame_length': 0.5625,
-            'frame_shift': 0.0625,
-            'high_freq': 4253,
-            'htk_compat': True,
-            'low_freq': 1367,
-            'num_mel_bins': 5,
-            'preemphasis_coefficient': 0.84,
-            'raw_energy': False,
-            'remove_dc_offset': True,
-            'round_to_power_of_two': True,
-            'snip_edges': True,
-            'subtract_mean': False,
-            'use_energy': True,
-            'use_log_fbank': True,
-            'use_power': False,
-            'vtln_high': 2112,
-            'vtln_low': 1445,
-            'vtln_warp': 1.0000,
-            'window_type': 'hamming',
-
-        }
         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.fbank(waveform, **kwargs)