[web] Update code from huggingface

runtime/web/app.py

@@ -1,22 +1,81 @@
-import json
+# Copyright (c) 2022 Horizon Robotics. (authors: Binbin Zhang)
+#               2022 Chengdong Liang (liangchengdong@mail.nwpu.edu.cn)
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 import gradio as gr
-import numpy as np
-import torch
-import wenetruntime as wenet
+import wenet
+
+# TODO: add hotword
+chs_model = wenet.load_model('chinese')
+en_model = wenet.load_model('english')
+
+
+def recognition(audio, lang='CN'):
+    if audio is None:
+        return "Input Error! Please enter one audio!"
+    # NOTE: model supports 16k sample_rate
+    if lang == 'CN':
+        ans = chs_model.transcribe(audio)
+    elif lang == 'EN':
+        ans = en_model.transcribe(audio)
+    else:
+        return "ERROR! Please select a language!"
+
+    if ans is None:
+        return "ERROR! No text output! Please try again!"
+    txt = ans['text']
+    return txt
 
-torch.manual_seed(777)  # for lint
 
-wenet.set_log_level(2)
-decoder = wenet.Decoder(lang='chs')
+# input
+inputs = [
+    gr.inputs.Audio(source="microphone", type="filepath", label='Input audio'),
+    gr.Radio(['EN', 'CN'], label='Language')
+]
 
-def recognition(audio):
-    sr, y = audio
-    assert sr in [48000, 16000]
-    if sr == 48000:  # Optional resample to 16000
-        y = (y / max(np.max(y), 1) * 32767)[::3].astype("int16")
-    ans = decoder.decode(y.tobytes(), True)
-    return json.loads(ans)
+output = gr.outputs.Textbox(label="Output Text")
+
+examples = [
+    ['examples/BAC009S0767W0127.wav', 'CN'],
+    ['examples/BAC009S0767W0424.wav', 'CN'],
+    ['examples/BAC009S0767W0488.wav', 'CN'],
+    ['examples/1995-1836-0002.flac', 'EN'],
+    ['examples/61-70968-0000.flac', 'EN'],
+    ['examples/672-122797-0000.flac', 'EN'],
+]
 
 text = "Speech Recognition in WeNet | 基于 WeNet 的语音识别"
-gr.Interface(recognition, inputs="mic", outputs="json",
-             description=text).launch()
+
+# description
+description = (
+    "Wenet Demo ! This is a speech recognition demo that supports Mandarin and English !"
+)
+
+article = (
+    "<p style='text-align: center'>"
+    "<a href='https://github.com/wenet-e2e/wenet' target='_blank'>Github: Learn more about WeNet</a>"
+    "</p>")
+
+interface = gr.Interface(
+    fn=recognition,
+    inputs=inputs,
+    outputs=output,
+    title=text,
+    description=description,
+    article=article,
+    examples=examples,
+    theme='huggingface',
+)
+
+interface.launch(enable_queue=True)

runtime/web/requirements.txt

@@ -1,2 +1,2 @@
-wenetruntime
-gradio
+wenet @ git+https://github.com/wenet-e2e/wenet.git
+gradio==3.14.0