Configurable low_freq high_freq, dithering (#664)

sherpa-onnx/python/csrc/features.cc

@@ -11,10 +11,17 @@ namespace sherpa_onnx {
 static void PybindFeatureExtractorConfig(py::module *m) {
   using PyClass = FeatureExtractorConfig;
   py::class_<PyClass>(*m, "FeatureExtractorConfig")
-      .def(py::init<int32_t, int32_t>(), py::arg("sampling_rate") = 16000,
-           py::arg("feature_dim") = 80)
+      .def(py::init<int32_t, int32_t, float, float, float>(),
+           py::arg("sampling_rate") = 16000,
+           py::arg("feature_dim") = 80,
+           py::arg("low_freq") = 20.0f,
+           py::arg("high_freq") = -400.0f,
+           py::arg("dither") = 0.0f)
       .def_readwrite("sampling_rate", &PyClass::sampling_rate)
       .def_readwrite("feature_dim", &PyClass::feature_dim)
+      .def_readwrite("low_freq", &PyClass::low_freq)
+      .def_readwrite("high_freq", &PyClass::high_freq)
+      .def_readwrite("dither", &PyClass::high_freq)
       .def("__str__", &PyClass::ToString);
 }
 

sherpa-onnx/python/sherpa_onnx/online_recognizer.py

@@ -41,6 +41,9 @@ class OnlineRecognizer(object):
         num_threads: int = 2,
         sample_rate: float = 16000,
         feature_dim: int = 80,
+        low_freq: float = 20.0,
+        high_freq: float = -400.0,
+        dither: float = 0.0,
         enable_endpoint_detection: bool = False,
         rule1_min_trailing_silence: float = 2.4,
         rule2_min_trailing_silence: float = 1.2,
@@ -80,6 +83,16 @@ class OnlineRecognizer(object):
             Sample rate of the training data used to train the model.
           feature_dim:
             Dimension of the feature used to train the model.
+          low_freq:
+            Low cutoff frequency for mel bins in feature extraction.
+          high_freq:
+            High cutoff frequency for mel bins in feature extraction
+            (if <= 0, offset from Nyquist)
+          dither:
+            Dithering constant (0.0 means no dither).
+            By default the audio samples are in range [-1,+1],
+            so dithering constant 0.00003 is a good value,
+            equivalent to the default 1.0 from kaldi
           enable_endpoint_detection:
             True to enable endpoint detection. False to disable endpoint
             detection.
@@ -140,6 +153,9 @@ class OnlineRecognizer(object):
         feat_config = FeatureExtractorConfig(
             sampling_rate=sample_rate,
             feature_dim=feature_dim,
+            low_freq=low_freq,
+            high_freq=high_freq,
+            dither=dither,
         )
 
         endpoint_config = EndpointConfig(