support google drive cloud service for future

app.py

@@ -21,6 +21,23 @@ import librosa.display
 import matplotlib.pyplot as plt
 
 
+# Google cloud service
+from googleapiclient.discovery import build
+from google.oauth2 import service_account
+from googleapiclient.http import MediaFileUpload
+import datetime
+
+# 来自Google Cloud控制台的JSON凭据文件
+credentials_file =  "./src/peerless-window-254907-b386b71c0d99.json"
+# "./client_secret_576367903492-diuopf97kn9eh1gte3vh65errtca1o64.apps.googleusercontent.com.json"
+# Google Drive API版本
+api_version = 'v3'
+
+# 创建服务对象
+credentials = service_account.Credentials.from_service_account_file(
+    credentials_file, scopes=['https://www.googleapis.com/auth/drive'])
+service = build('drive', api_version, credentials=credentials)
+
 # local import
 import sys
 
@@ -30,7 +47,6 @@ import lightning_module
 # Load automos
 # config_yaml = sys.argv[1]
 config_yaml = "config/Arthur.yaml"
-
 with open(config_yaml, "r") as f:
     # pdb.set_trace()
     try:
@@ -40,9 +56,9 @@ with open(config_yaml, "r") as f:
         exit()
 
 # Auto load examples
-
-with open(config["ref_txt"], "r") as f:
+with open(config['ref_txt'], "r") as f:
     refs = f.readlines()
+# refs = np.loadtxt(config["ref_txt"], delimiter="\n", dtype="str")
 refs_ids = [x.split()[0] for x in refs]
 refs_txt = [" ".join(x.split()[1:]) for x in refs]
 ref_feature = np.loadtxt(config["ref_feature"], delimiter=",", dtype="str")
@@ -112,7 +128,7 @@ class ChangeSampleRate(nn.Module):
 
 # MOS model
 model = lightning_module.BaselineLightningModule.load_from_checkpoint(
-    "./src/epoch=3-step=7459.ckpt"
+    "src/epoch=3-step=7459.ckpt"
 ).eval()
 
 # Get Speech Interval
@@ -138,11 +154,15 @@ def plot_UV(signal, audio_interv, sr):
     ax[1].set_ylim([-0.1, 1.1])
     return fig
 
+
+# Evaluation model
+
+
 def calc_mos(_, audio_path, id, ref, pre_ppm, fig=None):
     if audio_path == None:
         audio_path = _
         print("using ref audio as eval audio since it's empty")
-
+        
     wav, sr = torchaudio.load(audio_path)
     if wav.shape[0] != 1:
         wav = wav[0, :]
@@ -214,6 +234,9 @@ def calc_mos(_, audio_path, id, ref, pre_ppm, fig=None):
             "GOOD JOB! Please 【Save the Recording】.\nYou can start recording the next sample."
         )
 
+    # Google Drive saving # TODO
+    click_google_saving(audio_path)
+
     return (
         fig_h,
         predic_mos,
@@ -297,10 +320,11 @@ info = gr.Interface(
 if config["exp_id"] == None:
     config["exp_id"] = Path(config_yaml).stem
 
-## This is the theme for the interface
+## Theme
 css = """
 .ref_text textarea {font-size: 40px !important}
 .message textarea {font-size: 40px !important}
+
 """
 
 my_theme = gr.themes.Default().set(
@@ -313,6 +337,50 @@ my_theme = gr.themes.Default().set(
 # Callback for saving the recording
 callback = gr.CSVLogger()
 
+def generate_now_time_wav():
+    # Get the current date and time
+    current_time = datetime.datetime.now()
+
+    # Format the date and time as a string
+    time_string = current_time.strftime("%Y-%m-%d_%H-%M-%S")
+
+    # Create the WAV file name with the formatted time
+    wavfile_name = f"audio_{time_string}.wav"
+    return wavfile_name
+
+# Add google drive cloud saving
+def click_google_saving(audio_file,
+    
+                        ):
+                        # reference_id,
+                        # reference_textbox,
+                        # reference_PPM,
+                        # predict_mos,
+                        # hyp,
+                        # wer,
+                        # ppm,
+                        # msg,
+    name = generate_now_time_wav()
+    # 上传文件
+    media = MediaFileUpload(audio_file, mimetype='audio/wav')    
+    
+    request = service.files().create(
+        media_body=media,
+        body={'name': name, 
+              }
+    )
+            #   'reference_id': reference_id,
+            #   "reference_textbox": reference_textbox,
+            #   "reference_PPM": reference_PPM,
+            #   "predict_mos": predict_mos,
+            #   "hyp": hyp,
+            #   "wer": wer,
+            #   "ppm": ppm, 
+            #   "msg": msg
+    response = request.execute()
+    # return response.get('id')
+    
+
 with gr.Blocks(css=css, theme=my_theme) as demo:
     with gr.Column():
         with gr.Row():
@@ -450,6 +518,7 @@ with gr.Blocks(css=css, theme=my_theme) as demo:
                 preprocess=False,
                 api_name="flagging",
             )
+
         with gr.Row():
             b3 = gr.ClearButton(
                 [

local/check_data.py

@@ -0,0 +1,43 @@
+from googleapiclient.discovery import build
+from google.oauth2 import service_account
+from googleapiclient.http import MediaFileUpload
+import pdb
+pdb.set_trace()
+
+import gradio as gr
+
+# 来自Google Cloud控制台的JSON凭据文件
+credentials_file =  "./src/peerless-window-254907-b386b71c0d99.json"
+api_version = 'v3'
+
+# 创建服务对象
+credentials = service_account.Credentials.from_service_account_file(
+    credentials_file, scopes=['https://www.googleapis.com/auth/drive'])
+service = build('drive', api_version, credentials=credentials)
+
+# 列出文件
+results = service.files().list().execute()
+files = results.get('files', [])
+
+print(files)
+from googleapiclient.http import MediaIoBaseDownload
+import io 
+
+file_id = "1EqHciegNxZSyWJ9Nizo1QmRQEgTkgWCo"
+# Get the file's metadata
+file = service.files().get(fileId=file_id).execute()
+
+pdb.set_trace()
+request = service.files().get_media(fileId="1EqHciegNxZSyWJ9Nizo1QmRQEgTkgWCo")
+with open(file['name'], 'wb') as file_obj:
+    downloader = MediaIoBaseDownload(file_obj, request)
+    done = False
+    while not done:
+        status, done = downloader.next_chunk()
+        print(f"Download {int(status.progress() * 100)}%.")
+
+print(f"Downloaded: {file['name']}")
+
+pdb.set_trace()
+        
+# print('文件ID：%s' % response.get('id'))

local/test_google_drive.py

@@ -0,0 +1,93 @@
+from googleapiclient.discovery import build
+from google.oauth2 import service_account
+from googleapiclient.http import MediaFileUpload
+import pdb
+
+import gradio as gr
+
+# 来自Google Cloud控制台的JSON凭据文件
+credentials_file =  "./src/peerless-window-254907-b386b71c0d99.json"
+# "./client_secret_576367903492-diuopf97kn9eh1gte3vh65errtca1o64.apps.googleusercontent.com.json"
+# Google Drive API版本
+api_version = 'v3'
+
+# 创建服务对象
+credentials = service_account.Credentials.from_service_account_file(
+    credentials_file, scopes=['https://www.googleapis.com/auth/drive'])
+service = build('drive', api_version, credentials=credentials)
+
+
+import gradio as gr
+from transformers import pipeline
+import numpy as np
+import librosa
+import torchaudio
+
+import datetime
+
+def generate_now_time_wav():
+    # Get the current date and time
+    current_time = datetime.datetime.now()
+
+    # Format the date and time as a string
+    time_string = current_time.strftime("%Y-%m-%d_%H-%M-%S")
+
+    # Create the WAV file name with the formatted time
+    wavfile_name = f"audio_{time_string}.wav"
+    return wavfile_name
+
+# transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-base.en")
+
+def transcribe(audio_path):
+    if audio_path == None:
+        print("using ref audio as eval audio since it's empty")
+
+    wav, sr = torchaudio.load(audio_path)
+    if wav.shape[0] != 1:
+        wav = wav[0, :]
+    print(wav.shape)
+
+    name = generate_now_time_wav()
+    # 上传文件
+    media = MediaFileUpload(audio_path, mimetype='audio/wav')    
+    request = service.files().create(
+        media_body=media,
+        body={'name': name}
+    )
+    response = request.execute()
+    
+    return response.get('id')
+    
+demo = gr.Interface(
+    fn = transcribe,
+    inputs = gr.Audio(source="microphone", type='filepath'),
+    outputs = "text",
+)
+# file_path = 'data/3_michael_20230619_100/1st_session_ZOOM0015_002.wav'
+
+# x = gr.Audio(source="upload", type='filepath'),
+# pdb.set_trace()
+# x = transcribe(file_path)
+# pdb.set_trace()
+
+demo.launch()
+    
+# # 要上传的文件
+# file_name = '1st_session_ZOOM0015_001.wav'
+
+# # 上传文件
+# media = MediaFileUpload(file_path, mimetype='audio/wav')
+# request = service.files().create(
+#     media_body=media,
+#     body={'name': file_name}
+# )
+
+
+# response = request.execute()
+
+# # 列出文件
+# results = service.files().list().execute()
+# files = results.get('files', [])
+# pdb.set_trace()
+        
+# print('文件ID：%s' % response.get('id'))

src/peerless-window-254907-b386b71c0d99.json

@@ -0,0 +1,13 @@
+{
+  "type": "service_account",
+  "project_id": "peerless-window-254907",
+  "private_key_id": "b386b71c0d998879b5e47d776fba764d549a0696",
+  "private_key": "-----BEGIN PRIVATE KEY-----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\nIBCSOPWoWHqArBPZjzFDpoA=\n-----END PRIVATE KEY-----\n",
+  "client_email": "[email protected]",
+  "client_id": "100559289389957446034",
+  "auth_uri": "https://accounts.google.com/o/oauth2/auth",
+  "token_uri": "https://oauth2.googleapis.com/token",
+  "auth_provider_x509_cert_url": "https://www.googleapis.com/oauth2/v1/certs",
+  "client_x509_cert_url": "https://www.googleapis.com/robot/v1/metadata/x509/gradiotest%40peerless-window-254907.iam.gserviceaccount.com",
+  "universe_domain": "googleapis.com"
+}