Searched refs:window_step (Results 1 – 6 of 6) sorted by relevance
| /AliOS-Things-master/components/ai_agent/src/engine/tflite-micro/tensorflow/lite/experimental/microfrontend/python/kernel_tests/ |
| A D | audio_microfrontend_op_test.py | 50 window_step=WINDOW_STEP,
68 window_step=WINDOW_STEP,
89 window_step=WINDOW_STEP,
109 window_step=WINDOW_STEP,
131 window_step=WINDOW_STEP,
151 window_step=WINDOW_STEP,
|
| /AliOS-Things-master/components/ai_agent/src/engine/tflite-micro/tensorflow/lite/experimental/microfrontend/ops/ |
| A D | audio_microfrontend_op.cc | 74 int window_step; in __anon3fe811a90102() local
75 TF_RETURN_IF_ERROR(ctx->GetAttr("window_step", &window_step)); in __anon3fe811a90102()
76 window_step *= sample_rate / 1000; in __anon3fe811a90102()
93 ctx->Divide(num_frames, window_step, false, &num_frames)); in __anon3fe811a90102()
162 int window_step; in AudioMicrofrontendOp() local
163 OP_REQUIRES_OK(ctx, ctx->GetAttr("window_step", &window_step)); in AudioMicrofrontendOp()
164 config_.window.step_size_ms = window_step; in AudioMicrofrontendOp()
219 int window_step = config_.window.step_size_ms * sample_rate_ / 1000; in Compute() local
223 num_frames = (audio_size - window_size) / window_step + 1; in Compute()
|
| /AliOS-Things-master/components/ai_agent/src/engine/tflite-micro/tensorflow/lite/experimental/microfrontend/python/ops/ |
| A D | audio_microfrontend_op.py | 37 window_step=10, argument
109 audio, sample_rate, window_size, window_step, num_channels,
|
| /AliOS-Things-master/components/ai_agent/src/engine/tflite-micro/tensorflow/lite/experimental/microfrontend/ |
| A D | README.md | 48 `window_step=10ms`. The reason being that when reaching the point in the
55 audio with zeros such that we generate output frames based on the `window_step`.
|
| /AliOS-Things-master/solutions/tflite_micro_speech_demo/micro_speech/train/speech_commands/ |
| A D | freeze.py | 139 window_step=window_step_ms,
|
| A D | input_data.py | 477 window_step=window_step_ms,
|
Completed in 6 milliseconds