Spaces:
Running
Running
addupload_base_audio
Browse files- __pycache__/process.cpython-310.pyc +0 -0
- app.py +22 -133
- templates/index.html +75 -195
- templates/userRegister.html +225 -0
- transcription.py +26 -26
__pycache__/process.cpython-310.pyc
CHANGED
Binary files a/__pycache__/process.cpython-310.pyc and b/__pycache__/process.cpython-310.pyc differ
|
|
app.py
CHANGED
@@ -1,151 +1,34 @@
|
|
1 |
-
from flask import Flask, request, jsonify, render_template, send_from_directory
|
2 |
import base64
|
3 |
from pydub import AudioSegment # 変換用にpydubをインポート
|
4 |
import os
|
5 |
import shutil
|
6 |
-
|
7 |
-
import string
|
8 |
-
import random
|
9 |
-
from datetime import datetime, timedelta
|
10 |
-
from pyannote.audio import Model, Inference
|
11 |
-
from pydub import AudioSegment
|
12 |
-
from flask_sqlalchemy import SQLAlchemy
|
13 |
-
from dotenv import load_dotenv
|
14 |
-
from google.oauth2 import id_token
|
15 |
-
from google_auth_oauthlib.flow import Flow
|
16 |
-
from google.auth.transport import requests as google_requests
|
17 |
-
|
18 |
-
# Hugging Face のトークン取得(環境変数 HF に設定)
|
19 |
-
#hf_token = os.environ.get("HF")
|
20 |
-
load_dotenv()
|
21 |
-
hf_token = os.getenv("HF")
|
22 |
-
if hf_token is None:
|
23 |
-
raise ValueError("HUGGINGFACE_HUB_TOKEN が設定されていません。")
|
24 |
-
|
25 |
-
# キャッシュディレクトリの作成(書き込み可能な /tmp を利用)
|
26 |
-
cache_dir = "/tmp/hf_cache"
|
27 |
-
os.makedirs(cache_dir, exist_ok=True)
|
28 |
-
|
29 |
-
# pyannote モデルの読み込み
|
30 |
-
model = Model.from_pretrained("pyannote/embedding", use_auth_token=hf_token, cache_dir=cache_dir)
|
31 |
-
inference = Inference(model)
|
32 |
|
|
|
33 |
app = Flask(__name__)
|
34 |
|
35 |
-
app.config['SECRET_KEY'] = os.urandom(24)
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
def cosine_similarity(vec1, vec2):
|
41 |
-
vec1 = vec1 / np.linalg.norm(vec1)
|
42 |
-
vec2 = vec2 / np.linalg.norm(vec2)
|
43 |
-
return np.dot(vec1, vec2)
|
44 |
-
|
45 |
-
def segment_audio(path, target_path='/tmp/setup_voice', seg_duration=1.0):
|
46 |
-
"""
|
47 |
-
音声を指定秒数ごとに分割する。
|
48 |
-
target_path に分割したファイルを保存し、元の音声の総長(ミリ秒)を返す。
|
49 |
-
"""
|
50 |
-
os.makedirs(target_path, exist_ok=True)
|
51 |
-
base_sound = AudioSegment.from_file(path)
|
52 |
-
duration_ms = len(base_sound)
|
53 |
-
seg_duration_ms = int(seg_duration * 1000)
|
54 |
-
|
55 |
-
for i, start in enumerate(range(0, duration_ms, seg_duration_ms)):
|
56 |
-
end = min(start + seg_duration_ms, duration_ms)
|
57 |
-
segment = base_sound[start:end]
|
58 |
-
segment.export(os.path.join(target_path, f'{i}.wav'), format="wav")
|
59 |
-
|
60 |
-
return target_path, duration_ms
|
61 |
-
|
62 |
-
def calculate_similarity(path1, path2):
|
63 |
-
embedding1 = inference(path1)
|
64 |
-
embedding2 = inference(path2)
|
65 |
-
return float(cosine_similarity(embedding1.data.flatten(), embedding2.data.flatten()))
|
66 |
-
|
67 |
-
def process_audio(reference_path, input_path, output_folder='/tmp/data/matched_segments', seg_duration=1.0, threshold=0.5):
|
68 |
-
"""
|
69 |
-
入力音声ファイルを seg_duration 秒ごとに分割し、各セグメントと参照音声の類似度を計算。
|
70 |
-
類似度が threshold を超えたセグメントを output_folder にコピーし、マッチした時間(ms)と
|
71 |
-
マッチしなかった時間(ms)を返す。
|
72 |
-
"""
|
73 |
-
os.makedirs(output_folder, exist_ok=True)
|
74 |
-
segmented_path, total_duration_ms = segment_audio(input_path, seg_duration=seg_duration)
|
75 |
-
|
76 |
-
matched_time_ms = 0
|
77 |
-
for file in sorted(os.listdir(segmented_path)):
|
78 |
-
segment_file = os.path.join(segmented_path, file)
|
79 |
-
similarity = calculate_similarity(segment_file, reference_path)
|
80 |
-
if similarity > threshold:
|
81 |
-
shutil.copy(segment_file, output_folder)
|
82 |
-
matched_time_ms += len(AudioSegment.from_file(segment_file))
|
83 |
-
|
84 |
-
unmatched_time_ms = total_duration_ms - matched_time_ms
|
85 |
-
return matched_time_ms, unmatched_time_ms
|
86 |
-
|
87 |
-
def generate_random_string(length):
|
88 |
-
letters = string.ascii_letters + string.digits
|
89 |
-
return ''.join(random.choice(letters) for i in range(length))
|
90 |
-
|
91 |
-
def generate_filename(random_length):
|
92 |
-
random_string = generate_random_string(random_length)
|
93 |
-
current_time = datetime.now().strftime("%Y%m%d%H%M%S")
|
94 |
-
filename = f"{current_time}_{random_string}.wav"
|
95 |
-
return filename
|
96 |
-
|
97 |
# トップページ(テンプレート: index.html)
|
98 |
@app.route('/')
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
|
103 |
# フィードバック画面(テンプレート: feedback.html)
|
104 |
@app.route('/feedback', methods=['GET', 'POST'])
|
105 |
def feedback():
|
106 |
-
#ログイン問題解決しだい戻す
|
107 |
-
"""
|
108 |
-
if 'google_id' not in session:
|
109 |
-
return redirect(url_for('login'))
|
110 |
-
user_info = {
|
111 |
-
'name': session.get('name'),
|
112 |
-
'email': session.get('email')
|
113 |
-
}
|
114 |
-
"""
|
115 |
return render_template('feedback.html')
|
116 |
|
|
|
117 |
# 会話詳細画面(テンプレート: talkDetail.html)
|
118 |
@app.route('/talk_detail', methods=['GET', 'POST'])
|
119 |
def talk_detail():
|
120 |
-
"""
|
121 |
-
if 'google_id' not in session:
|
122 |
-
return redirect(url_for('login'))
|
123 |
-
user_info = {
|
124 |
-
'name': session.get('name'),
|
125 |
-
'email': session.get('email')
|
126 |
-
}
|
127 |
-
"""
|
128 |
return render_template('talkDetail.html')
|
129 |
|
130 |
-
|
131 |
-
@app.route('/
|
132 |
-
def
|
133 |
-
|
134 |
-
if 'google_id' not in session:
|
135 |
-
return redirect(url_for('login'))
|
136 |
-
user_info = {
|
137 |
-
'name': session.get('name'),
|
138 |
-
'email': session.get('email')
|
139 |
-
}
|
140 |
-
"""
|
141 |
-
return render_template('index.html')
|
142 |
-
|
143 |
-
@app.before_request
|
144 |
-
def before_request():
|
145 |
-
# リクエストのたびにセッションの寿命を更新する
|
146 |
-
session.permanent = True
|
147 |
-
app.permanent_session_lifetime = timedelta(minutes=15)
|
148 |
-
session.modified = True
|
149 |
|
150 |
# 音声アップロード&解析エンドポイント
|
151 |
@app.route('/upload_audio', methods=['POST'])
|
@@ -183,15 +66,23 @@ def upload_audio():
|
|
183 |
def upload_base_audio():
|
184 |
try:
|
185 |
data = request.get_json()
|
186 |
-
if not data or 'audio_data' not in data:
|
187 |
-
return jsonify({"error": "
|
188 |
|
189 |
# Base64デコードして音声バイナリを取得
|
190 |
audio_binary = base64.b64decode(data['audio_data'])
|
|
|
191 |
|
192 |
# 保存先ディレクトリの作成
|
193 |
audio_dir = "/tmp/data/base_audio"
|
194 |
os.makedirs(audio_dir, exist_ok=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
195 |
|
196 |
# 一時ファイルに保存(実際の形式は WebM などと仮定)
|
197 |
temp_audio_path = os.path.join(audio_dir, "temp_audio")
|
@@ -216,8 +107,6 @@ def upload_base_audio():
|
|
216 |
print("Error in /upload_base_audio:", str(e))
|
217 |
return jsonify({"error": "サーバーエラー", "details": str(e)}), 500
|
218 |
|
219 |
-
|
220 |
-
|
221 |
if __name__ == '__main__':
|
222 |
port = int(os.environ.get("PORT", 7860))
|
223 |
-
app.run(debug=True, host="0.0.0.0", port=port)
|
|
|
1 |
+
from flask import Flask, request, jsonify, render_template, send_from_directory
|
2 |
import base64
|
3 |
from pydub import AudioSegment # 変換用にpydubをインポート
|
4 |
import os
|
5 |
import shutil
|
6 |
+
from process import AudioProcessor
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
|
8 |
+
process=AudioProcessor()
|
9 |
app = Flask(__name__)
|
10 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
# トップページ(テンプレート: index.html)
|
12 |
@app.route('/')
|
13 |
+
@app.route('/index', methods=['GET', 'POST'])
|
14 |
+
def index():
|
15 |
+
return render_template('index.html')
|
16 |
|
17 |
# フィードバック画面(テンプレート: feedback.html)
|
18 |
@app.route('/feedback', methods=['GET', 'POST'])
|
19 |
def feedback():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
return render_template('feedback.html')
|
21 |
|
22 |
+
|
23 |
# 会話詳細画面(テンプレート: talkDetail.html)
|
24 |
@app.route('/talk_detail', methods=['GET', 'POST'])
|
25 |
def talk_detail():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
return render_template('talkDetail.html')
|
27 |
|
28 |
+
# 音声登録画面(テンプレート: userRegister.html)
|
29 |
+
@app.route('/userregister', methods=['GET', 'POST'])
|
30 |
+
def userregister():
|
31 |
+
return render_template('userregister.html')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
|
33 |
# 音声アップロード&解析エンドポイント
|
34 |
@app.route('/upload_audio', methods=['POST'])
|
|
|
66 |
def upload_base_audio():
|
67 |
try:
|
68 |
data = request.get_json()
|
69 |
+
if not data or 'audio_data' not in data or 'name' not in data:
|
70 |
+
return jsonify({"error": "音声データまたは名前がありません"}), 400
|
71 |
|
72 |
# Base64デコードして音声バイナリを取得
|
73 |
audio_binary = base64.b64decode(data['audio_data'])
|
74 |
+
name = data['name'] # 名前を取得
|
75 |
|
76 |
# 保存先ディレクトリの作成
|
77 |
audio_dir = "/tmp/data/base_audio"
|
78 |
os.makedirs(audio_dir, exist_ok=True)
|
79 |
+
|
80 |
+
# 辞書型を作成(音声データと名前)
|
81 |
+
audio_info = {
|
82 |
+
"name": name,
|
83 |
+
"audio_data": audio_binary # バイナリデータをそのまま格納
|
84 |
+
}
|
85 |
+
|
86 |
|
87 |
# 一時ファイルに保存(実際の形式は WebM などと仮定)
|
88 |
temp_audio_path = os.path.join(audio_dir, "temp_audio")
|
|
|
107 |
print("Error in /upload_base_audio:", str(e))
|
108 |
return jsonify({"error": "サーバーエラー", "details": str(e)}), 500
|
109 |
|
|
|
|
|
110 |
if __name__ == '__main__':
|
111 |
port = int(os.environ.get("PORT", 7860))
|
112 |
+
app.run(debug=True, host="0.0.0.0", port=port)
|
templates/index.html
CHANGED
@@ -4,80 +4,52 @@
|
|
4 |
<meta charset="UTF-8" />
|
5 |
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
|
6 |
<title>Voice Recorder Interface</title>
|
7 |
-
<
|
8 |
-
|
9 |
-
|
10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
<style>
|
12 |
-
body {
|
13 |
-
display: flex;
|
14 |
-
flex-direction: column;
|
15 |
-
justify-content: center;
|
16 |
-
align-items: center;
|
17 |
-
height: 100vh;
|
18 |
-
margin: 0;
|
19 |
-
background-color: #121212;
|
20 |
-
color: white;
|
21 |
-
}
|
22 |
-
/* トグルスイッチ(基準音声保存用) */
|
23 |
-
.toggle-container {
|
24 |
-
display: flex;
|
25 |
-
align-items: center;
|
26 |
-
margin-bottom: 20px;
|
27 |
-
}
|
28 |
-
.toggle-label {
|
29 |
-
margin-right: 10px;
|
30 |
-
}
|
31 |
-
.toggle-switch {
|
32 |
-
position: relative;
|
33 |
-
display: inline-block;
|
34 |
-
width: 50px;
|
35 |
-
height: 24px;
|
36 |
-
}
|
37 |
-
.toggle-switch input {
|
38 |
-
opacity: 0;
|
39 |
-
width: 0;
|
40 |
-
height: 0;
|
41 |
-
}
|
42 |
-
.slider {
|
43 |
-
position: absolute;
|
44 |
-
cursor: pointer;
|
45 |
-
top: 0;
|
46 |
-
left: 0;
|
47 |
-
right: 0;
|
48 |
-
bottom: 0;
|
49 |
-
background-color: #757575;
|
50 |
-
transition: 0.2s;
|
51 |
-
border-radius: 34px;
|
52 |
-
}
|
53 |
-
.slider::before {
|
54 |
-
content: "";
|
55 |
-
position: absolute;
|
56 |
-
height: 18px;
|
57 |
-
width: 18px;
|
58 |
-
left: 4px;
|
59 |
-
bottom: 3px;
|
60 |
-
background-color: white;
|
61 |
-
transition: 0.2s;
|
62 |
-
border-radius: 50%;
|
63 |
-
}
|
64 |
-
input:checked + .slider {
|
65 |
-
background-color: #4caf50;
|
66 |
-
}
|
67 |
-
input:checked + .slider::before {
|
68 |
-
transform: translateX(26px);
|
69 |
-
}
|
70 |
-
/* チャートのスタイル */
|
71 |
-
.chart {
|
72 |
-
width: 300px;
|
73 |
-
height: 300px;
|
74 |
-
margin-bottom: 20px; /* 円グラフとボタンの間隔を狭く */
|
75 |
-
}
|
76 |
-
.controls {
|
77 |
-
display: flex;
|
78 |
-
flex-direction: column;
|
79 |
-
align-items: center;
|
80 |
-
}
|
81 |
.record-button {
|
82 |
width: 80px;
|
83 |
height: 80px;
|
@@ -91,6 +63,7 @@
|
|
91 |
box-shadow: 0 4px 6px rgba(0, 0, 0, 0.4);
|
92 |
transition: all 0.2s ease;
|
93 |
}
|
|
|
94 |
.record-icon {
|
95 |
width: 60px;
|
96 |
height: 60px;
|
@@ -98,83 +71,35 @@
|
|
98 |
border-radius: 50%;
|
99 |
transition: all 0.2s ease;
|
100 |
}
|
|
|
101 |
.recording .record-icon {
|
102 |
width: 40px;
|
103 |
height: 40px;
|
104 |
border-radius: 10%;
|
105 |
}
|
106 |
-
.result-button {
|
107 |
-
margin-left: 10px;
|
108 |
-
|
109 |
-
margin-top: 20px;
|
110 |
-
padding: 10px 20px;
|
111 |
-
background-color: #4caf50;
|
112 |
-
border: none;
|
113 |
-
border-radius: 5px;
|
114 |
-
color: white;
|
115 |
-
cursor: pointer;
|
116 |
-
box-shadow: 0 4px 6px rgba(0, 0, 0, 0.4);
|
117 |
-
}
|
118 |
-
.result {
|
119 |
-
display: flex;
|
120 |
-
}
|
121 |
-
.result-button:hover {
|
122 |
-
background-color: #388e3c;
|
123 |
-
}
|
124 |
-
header {
|
125 |
-
display: flex;
|
126 |
-
}
|
127 |
</style>
|
128 |
-
<script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
|
129 |
-
</head>
|
130 |
-
<body>
|
131 |
-
<!-- トグルスイッチ:基準音声保存モード -->
|
132 |
-
<div class="toggle-container">
|
133 |
-
<span class="toggle-label">基準音声を保存</span>
|
134 |
-
<label class="toggle-switch">
|
135 |
-
<input type="checkbox" id="baseVoiceToggle" />
|
136 |
-
<span class="slider"></span>
|
137 |
-
</label>
|
138 |
-
</div>
|
139 |
-
|
140 |
-
<!-- チャート表示部 -->
|
141 |
-
<div class="chart">
|
142 |
-
<canvas id="speechChart"></canvas>
|
143 |
-
</div>
|
144 |
-
|
145 |
-
<!-- 録音ボタン -->
|
146 |
-
<button class="record-button" id="recordButton" onclick="toggleRecording()">
|
147 |
-
<div class="record-icon" id="recordIcon"></div>
|
148 |
-
</button>
|
149 |
-
|
150 |
-
<!-- 結果ボタン -->
|
151 |
-
<div class="result-buttons">
|
152 |
-
<button class="result-button" id="historyButton" onclick="showHistory()">
|
153 |
-
会話履歴を表示
|
154 |
-
</button>
|
155 |
-
<button class="result-button" id="feedbackButton" onclick="showResults()">
|
156 |
-
フィードバック画面を表示
|
157 |
-
</button>
|
158 |
-
</div>
|
159 |
|
160 |
<script>
|
161 |
let isRecording = false;
|
162 |
let mediaRecorder;
|
163 |
let audioChunks = [];
|
164 |
-
let recordingInterval;
|
165 |
-
let baseTimeout; // 基準音声モード用のタイマー
|
166 |
let count_voice = 0;
|
167 |
let before_rate = 0;
|
168 |
|
|
|
|
|
|
|
|
|
169 |
// Chart.js の初期化
|
170 |
const ctx = document.getElementById("speechChart").getContext("2d");
|
171 |
const speechChart = new Chart(ctx, {
|
172 |
type: "doughnut",
|
173 |
data: {
|
174 |
-
labels:
|
175 |
datasets: [
|
176 |
{
|
177 |
-
data:
|
178 |
backgroundColor: ["#4caf50", "#757575"],
|
179 |
},
|
180 |
],
|
@@ -191,9 +116,20 @@
|
|
191 |
},
|
192 |
});
|
193 |
|
194 |
-
|
195 |
-
function
|
196 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
197 |
}
|
198 |
|
199 |
async function toggleRecording() {
|
@@ -222,26 +158,6 @@
|
|
222 |
};
|
223 |
|
224 |
mediaRecorder.start();
|
225 |
-
|
226 |
-
if (isBaseVoiceMode()) {
|
227 |
-
// 基準音声モード:10秒後に自動停止するタイマーをセット
|
228 |
-
baseTimeout = setTimeout(() => {
|
229 |
-
if (mediaRecorder && mediaRecorder.state === "recording") {
|
230 |
-
mediaRecorder.stop();
|
231 |
-
// 10秒経過しても録音ボタンがONなら強制的に停止&トグルをオフにする
|
232 |
-
isRecording = false;
|
233 |
-
recordButton.classList.remove("recording");
|
234 |
-
document.getElementById("baseVoiceToggle").checked = false;
|
235 |
-
}
|
236 |
-
}, 10000);
|
237 |
-
} else {
|
238 |
-
// 通常モード:10秒ごとに自動停止して送信、継続録音する処理
|
239 |
-
recordingInterval = setInterval(() => {
|
240 |
-
if (mediaRecorder && mediaRecorder.state === "recording") {
|
241 |
-
mediaRecorder.stop();
|
242 |
-
}
|
243 |
-
}, 10000);
|
244 |
-
}
|
245 |
} catch (error) {
|
246 |
console.error("マイクへのアクセスに失敗しました:", error);
|
247 |
isRecording = false;
|
@@ -251,11 +167,6 @@
|
|
251 |
// 手動停止
|
252 |
isRecording = false;
|
253 |
recordButton.classList.remove("recording");
|
254 |
-
if (isBaseVoiceMode()) {
|
255 |
-
clearTimeout(baseTimeout);
|
256 |
-
} else {
|
257 |
-
clearInterval(recordingInterval);
|
258 |
-
}
|
259 |
if (mediaRecorder && mediaRecorder.state === "recording") {
|
260 |
mediaRecorder.stop();
|
261 |
count_voice = 0;
|
@@ -269,11 +180,7 @@
|
|
269 |
const reader = new FileReader();
|
270 |
reader.onloadend = () => {
|
271 |
const base64String = reader.result.split(",")[1]; // Base64エンコードされた音声データ
|
272 |
-
|
273 |
-
const endpoint = isBaseVoiceMode()
|
274 |
-
? "/upload_base_audio"
|
275 |
-
: "/upload_audio";
|
276 |
-
fetch(endpoint, {
|
277 |
method: "POST",
|
278 |
headers: {
|
279 |
"Content-Type": "application/json",
|
@@ -285,7 +192,7 @@
|
|
285 |
if (data.error) {
|
286 |
alert("エラー: " + data.error);
|
287 |
console.error(data.details);
|
288 |
-
} else if (data.rate !== undefined
|
289 |
// 通常モードの場合、解析結果をチャートに反映
|
290 |
if (count_voice === 0) {
|
291 |
speechChart.data.datasets[0].data = [
|
@@ -310,44 +217,17 @@
|
|
310 |
}
|
311 |
count_voice++;
|
312 |
speechChart.update();
|
313 |
-
} else {
|
314 |
-
// 基準音声モードまたは解析結果がない場合
|
315 |
-
if (isBaseVoiceMode()) {
|
316 |
-
//alert('基準音声が保存されました。');
|
317 |
-
// トグルをリセット
|
318 |
-
document.getElementById("baseVoiceToggle").checked = false;
|
319 |
-
} else {
|
320 |
-
//alert('音声がバックエンドに送信されました。');
|
321 |
-
}
|
322 |
-
}
|
323 |
-
// 通常モードの場合、録音が継続中なら次の録音を開始(自動連続録音)
|
324 |
-
if (
|
325 |
-
!isBaseVoiceMode() &&
|
326 |
-
isRecording &&
|
327 |
-
mediaRecorder &&
|
328 |
-
mediaRecorder.state === "inactive"
|
329 |
-
) {
|
330 |
-
mediaRecorder.start();
|
331 |
}
|
332 |
})
|
333 |
.catch((error) => {
|
334 |
console.error("エラー:", error);
|
335 |
-
if (
|
336 |
-
!isBaseVoiceMode() &&
|
337 |
-
isRecording &&
|
338 |
-
mediaRecorder &&
|
339 |
-
mediaRecorder.state === "inactive"
|
340 |
-
) {
|
341 |
-
mediaRecorder.start();
|
342 |
-
}
|
343 |
});
|
344 |
};
|
345 |
reader.readAsDataURL(audioBlob);
|
346 |
}
|
347 |
|
348 |
-
function
|
349 |
-
window.location.href = "
|
350 |
-
alert("会話履歴を表示する機能は未実装です。");
|
351 |
}
|
352 |
|
353 |
function showResults() {
|
@@ -355,9 +235,9 @@
|
|
355 |
window.location.href = "feedback";
|
356 |
}
|
357 |
|
358 |
-
function
|
359 |
-
//
|
360 |
-
window.location.href = "
|
361 |
}
|
362 |
</script>
|
363 |
</body>
|
|
|
4 |
<meta charset="UTF-8" />
|
5 |
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
|
6 |
<title>Voice Recorder Interface</title>
|
7 |
+
<script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
|
8 |
+
<script src="https://cdn.tailwindcss.com"></script>
|
9 |
+
</head>
|
10 |
+
<body
|
11 |
+
class="flex flex-col items-center justify-center h-screen bg-gray-900 text-white"
|
12 |
+
>
|
13 |
+
<!-- メンバーを登録ボタン -->
|
14 |
+
<div class="flex items-center mb-5">
|
15 |
+
<button
|
16 |
+
id="registerButton"
|
17 |
+
onclick="showUserRegister()"
|
18 |
+
class="px-4 py-2 bg-blue-600 rounded-md hover:bg-blue-700 transition"
|
19 |
+
>
|
20 |
+
メンバーを登録
|
21 |
+
</button>
|
22 |
+
</div>
|
23 |
+
|
24 |
+
<!-- チャート表示部 -->
|
25 |
+
<div class="chart w-72 h-72 mb-5">
|
26 |
+
<canvas id="speechChart"></canvas>
|
27 |
+
</div>
|
28 |
+
|
29 |
+
<!-- 録音ボタン -->
|
30 |
+
<button class="record-button" id="recordButton" onclick="toggleRecording()">
|
31 |
+
<div class="record-icon" id="recordIcon"></div>
|
32 |
+
</button>
|
33 |
+
|
34 |
+
<!-- 結果ボタン -->
|
35 |
+
<div class="flex mt-5">
|
36 |
+
<button
|
37 |
+
id="historyButton"
|
38 |
+
onclick="showTalkdetail()"
|
39 |
+
class="result-button px-4 py-2 mx-2 bg-green-600 rounded-md hover:bg-green-700 transition"
|
40 |
+
>
|
41 |
+
会話履歴を表示
|
42 |
+
</button>
|
43 |
+
<button
|
44 |
+
id="feedbackButton"
|
45 |
+
onclick="showResults()"
|
46 |
+
class="result-button px-4 py-2 mx-2 bg-blue-600 rounded-md hover:bg-blue-700 transition"
|
47 |
+
>
|
48 |
+
フィードバック画面を表示
|
49 |
+
</button>
|
50 |
+
</div>
|
51 |
+
|
52 |
<style>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
.record-button {
|
54 |
width: 80px;
|
55 |
height: 80px;
|
|
|
63 |
box-shadow: 0 4px 6px rgba(0, 0, 0, 0.4);
|
64 |
transition: all 0.2s ease;
|
65 |
}
|
66 |
+
|
67 |
.record-icon {
|
68 |
width: 60px;
|
69 |
height: 60px;
|
|
|
71 |
border-radius: 50%;
|
72 |
transition: all 0.2s ease;
|
73 |
}
|
74 |
+
|
75 |
.recording .record-icon {
|
76 |
width: 40px;
|
77 |
height: 40px;
|
78 |
border-radius: 10%;
|
79 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
80 |
</style>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
81 |
|
82 |
<script>
|
83 |
let isRecording = false;
|
84 |
let mediaRecorder;
|
85 |
let audioChunks = [];
|
86 |
+
let recordingInterval;
|
|
|
87 |
let count_voice = 0;
|
88 |
let before_rate = 0;
|
89 |
|
90 |
+
// 初期設定:人数と名前を受け取って円グラフを作成
|
91 |
+
let members = ["自分", "ahaha", "nufufu", "不明"];
|
92 |
+
let voiceData = [50, 20, 20, 10]; // 自分と不明の割合を仮設定
|
93 |
+
|
94 |
// Chart.js の初期化
|
95 |
const ctx = document.getElementById("speechChart").getContext("2d");
|
96 |
const speechChart = new Chart(ctx, {
|
97 |
type: "doughnut",
|
98 |
data: {
|
99 |
+
labels: members,
|
100 |
datasets: [
|
101 |
{
|
102 |
+
data: voiceData,
|
103 |
backgroundColor: ["#4caf50", "#757575"],
|
104 |
},
|
105 |
],
|
|
|
116 |
},
|
117 |
});
|
118 |
|
119 |
+
//録音ボタン見た目変化
|
120 |
+
function toggleRecording() {
|
121 |
+
isRecording = !isRecording;
|
122 |
+
const recordIcon = document.getElementById("recordIcon");
|
123 |
+
if (isRecording) {
|
124 |
+
recordIcon.classList.add("w-10", "h-10", "bg-red-900", "rounded-md");
|
125 |
+
} else {
|
126 |
+
recordIcon.classList.remove(
|
127 |
+
"w-10",
|
128 |
+
"h-10",
|
129 |
+
"bg-red-900",
|
130 |
+
"rounded-md"
|
131 |
+
);
|
132 |
+
}
|
133 |
}
|
134 |
|
135 |
async function toggleRecording() {
|
|
|
158 |
};
|
159 |
|
160 |
mediaRecorder.start();
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
161 |
} catch (error) {
|
162 |
console.error("マイクへのアクセスに失敗しました:", error);
|
163 |
isRecording = false;
|
|
|
167 |
// 手動停止
|
168 |
isRecording = false;
|
169 |
recordButton.classList.remove("recording");
|
|
|
|
|
|
|
|
|
|
|
170 |
if (mediaRecorder && mediaRecorder.state === "recording") {
|
171 |
mediaRecorder.stop();
|
172 |
count_voice = 0;
|
|
|
180 |
const reader = new FileReader();
|
181 |
reader.onloadend = () => {
|
182 |
const base64String = reader.result.split(",")[1]; // Base64エンコードされた音声データ
|
183 |
+
fetch("/upload_audio", {
|
|
|
|
|
|
|
|
|
184 |
method: "POST",
|
185 |
headers: {
|
186 |
"Content-Type": "application/json",
|
|
|
192 |
if (data.error) {
|
193 |
alert("エラー: " + data.error);
|
194 |
console.error(data.details);
|
195 |
+
} else if (data.rate !== undefined) {
|
196 |
// 通常モードの場合、解析結果をチャートに反映
|
197 |
if (count_voice === 0) {
|
198 |
speechChart.data.datasets[0].data = [
|
|
|
217 |
}
|
218 |
count_voice++;
|
219 |
speechChart.update();
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
220 |
}
|
221 |
})
|
222 |
.catch((error) => {
|
223 |
console.error("エラー:", error);
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
224 |
});
|
225 |
};
|
226 |
reader.readAsDataURL(audioBlob);
|
227 |
}
|
228 |
|
229 |
+
function showTalkdetail() {
|
230 |
+
window.location.href = "talk_detail";
|
|
|
231 |
}
|
232 |
|
233 |
function showResults() {
|
|
|
235 |
window.location.href = "feedback";
|
236 |
}
|
237 |
|
238 |
+
function showUserRegister() {
|
239 |
+
// 音声登録画面へ遷移
|
240 |
+
window.location.href = "userregister";
|
241 |
}
|
242 |
</script>
|
243 |
</body>
|
templates/userRegister.html
ADDED
@@ -0,0 +1,225 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
<!DOCTYPE html>
|
2 |
+
<html lang="ja">
|
3 |
+
<head>
|
4 |
+
<meta charset="UTF-8" />
|
5 |
+
<title>ユーザー音声登録</title>
|
6 |
+
<script src="https://cdn.tailwindcss.com"></script>
|
7 |
+
<style>
|
8 |
+
@keyframes pulse-scale {
|
9 |
+
0%,
|
10 |
+
100% {
|
11 |
+
transform: scale(1);
|
12 |
+
}
|
13 |
+
50% {
|
14 |
+
transform: scale(1.1);
|
15 |
+
}
|
16 |
+
}
|
17 |
+
.animate-pulse-scale {
|
18 |
+
animation: pulse-scale 1s infinite;
|
19 |
+
}
|
20 |
+
.record-button {
|
21 |
+
width: 50px;
|
22 |
+
height: 50px;
|
23 |
+
background-color: transparent;
|
24 |
+
border-radius: 50%;
|
25 |
+
border: 2px solid white;
|
26 |
+
display: flex;
|
27 |
+
justify-content: center;
|
28 |
+
align-items: center;
|
29 |
+
cursor: pointer;
|
30 |
+
box-shadow: 0 4px 6px rgba(0, 0, 0, 0.4);
|
31 |
+
transition: all 0.3s ease;
|
32 |
+
}
|
33 |
+
.record-icon {
|
34 |
+
width: 35px;
|
35 |
+
height: 35px;
|
36 |
+
background-color: #d32f2f;
|
37 |
+
border-radius: 50%;
|
38 |
+
transition: all 0.3s ease;
|
39 |
+
}
|
40 |
+
.record-button.recording .record-icon {
|
41 |
+
border-radius: 4px; /* 録音時に赤い部分だけ四角にする */
|
42 |
+
}
|
43 |
+
.recording .record-icon {
|
44 |
+
width: 20px;
|
45 |
+
height: 20px;
|
46 |
+
border-radius: 50%;
|
47 |
+
}
|
48 |
+
@media (max-width: 640px) {
|
49 |
+
.container {
|
50 |
+
padding: 2rem;
|
51 |
+
}
|
52 |
+
}
|
53 |
+
</style>
|
54 |
+
</head>
|
55 |
+
<body
|
56 |
+
class="bg-gray-800 text-gray-100 dark:bg-gray-900 dark:text-gray-300 transition-colors"
|
57 |
+
>
|
58 |
+
<div class="container mx-auto p-5 max-w-full sm:max-w-2xl">
|
59 |
+
<div id="people-list" class="space-y-4"></div>
|
60 |
+
<button
|
61 |
+
id="add-btn"
|
62 |
+
class="mt-6 px-6 py-2 bg-green-500 text-white rounded-lg hover:bg-green-600 transition-colors"
|
63 |
+
>
|
64 |
+
メンバーを追加
|
65 |
+
</button>
|
66 |
+
|
67 |
+
<!-- 録音画面に戻るボタン -->
|
68 |
+
<button
|
69 |
+
id="backButton"
|
70 |
+
class="mt-6 px-6 py-2 bg-blue-500 text-white rounded-lg hover:bg-blue-600 transition-colors"
|
71 |
+
>
|
72 |
+
録音画面に戻る
|
73 |
+
</button>
|
74 |
+
</div>
|
75 |
+
|
76 |
+
<script>
|
77 |
+
let mediaRecorder;
|
78 |
+
let audioChunks = [];
|
79 |
+
let userCount = 0; // 追加されたメンバー数を保持
|
80 |
+
let isRecording = false; // 録音中かどうかを判定するフラグ
|
81 |
+
let currentRecordingButton = null; // 現在録音中のボタンを保持
|
82 |
+
|
83 |
+
function toggleRecording(button) {
|
84 |
+
button.classList.toggle("recording");
|
85 |
+
}
|
86 |
+
|
87 |
+
async function startRecording(button) {
|
88 |
+
if (isRecording && currentRecordingButton !== button) return; // 他の人が録音中なら何もしない
|
89 |
+
isRecording = true; // 録音中に設定
|
90 |
+
currentRecordingButton = button; // 録音中のボタンを記録
|
91 |
+
|
92 |
+
try {
|
93 |
+
const stream = await navigator.mediaDevices.getUserMedia({
|
94 |
+
audio: true,
|
95 |
+
});
|
96 |
+
mediaRecorder = new MediaRecorder(stream, { mimeType: "audio/webm" });
|
97 |
+
audioChunks = [];
|
98 |
+
mediaRecorder.ondataavailable = (e) => audioChunks.push(e.data);
|
99 |
+
mediaRecorder.onstop = () => {
|
100 |
+
sendAudioChunks(audioChunks, button); // ボタン情報を渡す
|
101 |
+
audioChunks = [];
|
102 |
+
isRecording = false; // 録音停止後はフラグを戻す
|
103 |
+
currentRecordingButton = null; // 録音ボタンを解除
|
104 |
+
};
|
105 |
+
mediaRecorder.start();
|
106 |
+
toggleRecording(button);
|
107 |
+
} catch (err) {
|
108 |
+
console.error("マイクアクセスに失敗しました:", err);
|
109 |
+
isRecording = false; // エラー発生時もフラグを戻す
|
110 |
+
currentRecordingButton = null;
|
111 |
+
}
|
112 |
+
}
|
113 |
+
|
114 |
+
function stopRecording(button) {
|
115 |
+
if (!isRecording) return; // 録音中でない場合は停止しない
|
116 |
+
mediaRecorder.stop();
|
117 |
+
toggleRecording(button);
|
118 |
+
}
|
119 |
+
|
120 |
+
function handleRecording(e) {
|
121 |
+
const button = e.target.closest(".record-button");
|
122 |
+
if (button) {
|
123 |
+
if (isRecording && currentRecordingButton !== button) {
|
124 |
+
// 他の人が録音中なら反応しない
|
125 |
+
return;
|
126 |
+
}
|
127 |
+
if (mediaRecorder && mediaRecorder.state === "recording") {
|
128 |
+
stopRecording(button);
|
129 |
+
} else {
|
130 |
+
startRecording(button);
|
131 |
+
}
|
132 |
+
}
|
133 |
+
}
|
134 |
+
|
135 |
+
function sendAudioChunks(chunks, button) {
|
136 |
+
// 引数に button を追加
|
137 |
+
const audioBlob = new Blob(chunks, { type: "audio/wav" });
|
138 |
+
const reader = new FileReader();
|
139 |
+
reader.onloadend = () => {
|
140 |
+
const base64String = reader.result.split(",")[1]; // Base64エンコードされた音声データ
|
141 |
+
const form = button.closest("form");
|
142 |
+
const nameInput = form.querySelector('input[name="name"]');
|
143 |
+
const name = nameInput ? nameInput.value : "unknown"; // 名前がない
|
144 |
+
fetch("/upload_base_audio", {
|
145 |
+
method: "POST",
|
146 |
+
headers: {
|
147 |
+
"Content-Type": "application/json",
|
148 |
+
},
|
149 |
+
body: JSON.stringify({ audio_data: base64String, name: name }),
|
150 |
+
})
|
151 |
+
.then((response) => response.json())
|
152 |
+
.then((data) => {
|
153 |
+
// エラー処理のみ残す
|
154 |
+
if (data.error) {
|
155 |
+
alert("エラー: " + data.error);
|
156 |
+
console.error(data.details);
|
157 |
+
}
|
158 |
+
// 成功時の処理(ボタンの有効化など)
|
159 |
+
else {
|
160 |
+
console.log("音声データ送信成功:", data);
|
161 |
+
// 必要に応じて、ここでUIの変更(ボタンの有効化など)を行う
|
162 |
+
// 例: button.disabled = true; // 送信ボタンを無効化
|
163 |
+
// 例: button.classList.remove("recording"); //録音中のスタイルを解除
|
164 |
+
}
|
165 |
+
})
|
166 |
+
.catch((error) => {
|
167 |
+
console.error("エラー:", error);
|
168 |
+
});
|
169 |
+
};
|
170 |
+
reader.readAsDataURL(audioBlob);
|
171 |
+
}
|
172 |
+
|
173 |
+
document.getElementById("add-btn").addEventListener("click", () => {
|
174 |
+
const newItem = document.createElement("div");
|
175 |
+
newItem.className = "flex items-center gap-3 flex-wrap";
|
176 |
+
newItem.innerHTML = `
|
177 |
+
<form
|
178 |
+
action="/submit"
|
179 |
+
method="POST"
|
180 |
+
class="flex items-center space-x-2 w-full sm:w-auto"
|
181 |
+
onsubmit="event.preventDefault();"
|
182 |
+
>
|
183 |
+
<input
|
184 |
+
type="text"
|
185 |
+
name="name"
|
186 |
+
placeholder="名前を入力"
|
187 |
+
class="flex-1 px-4 py-2 border rounded-lg focus:outline-none focus:ring-2 focus:ring-blue-500 bg-gray-700 text-white"
|
188 |
+
/>
|
189 |
+
|
190 |
+
<button type="button" class="record-button" aria-label="音声録音開始">
|
191 |
+
<div class="record-icon"></div>
|
192 |
+
</button>
|
193 |
+
|
194 |
+
<button
|
195 |
+
type="submit"
|
196 |
+
class="submit-button px-4 py-2 border rounded-lg bg-blue-500 text-white hover:bg-blue-600"
|
197 |
+
>
|
198 |
+
送信
|
199 |
+
</button>
|
200 |
+
</form>
|
201 |
+
`;
|
202 |
+
newItem.addEventListener("click", handleRecording);
|
203 |
+
document.getElementById("people-list").appendChild(newItem);
|
204 |
+
userCount++; // 新しいメンバーを追加するたびにカウントを増やす
|
205 |
+
});
|
206 |
+
|
207 |
+
// 「録音画面に戻る」ボタンの処理
|
208 |
+
document
|
209 |
+
.getElementById("backButton")
|
210 |
+
.addEventListener("click", function () {
|
211 |
+
// メンバーの人数を送信する
|
212 |
+
sendUserCount();
|
213 |
+
|
214 |
+
// index.htmlに戻る
|
215 |
+
window.location.href = "index.html";
|
216 |
+
});
|
217 |
+
|
218 |
+
// メンバーの人数を送信する関数
|
219 |
+
function sendUserCount() {
|
220 |
+
console.log(`追加された人数: ${userCount}`);
|
221 |
+
// ここで人数を送信する処理を実行(例: fetchを使ってサーバーに送信)
|
222 |
+
}
|
223 |
+
</script>
|
224 |
+
</body>
|
225 |
+
</html>
|
transcription.py
CHANGED
@@ -2,7 +2,7 @@ import os
|
|
2 |
from faster_whisper import WhisperModel
|
3 |
|
4 |
class TranscriptionMaker():
|
5 |
-
|
6 |
def __init__(self,output_dir=os.path.abspath("/tmp/data/transcriptions")):
|
7 |
self.model = WhisperModel("base", device="cpu")
|
8 |
self.output_dir = output_dir
|
@@ -13,35 +13,35 @@ class TranscriptionMaker():
|
|
13 |
print(f"Error creating directory {self.output_dir}: {e}")
|
14 |
raise
|
15 |
|
16 |
-
|
17 |
-
def create_transcription(self,
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
for segment in segments:
|
26 |
results.append({
|
27 |
"start": segment.start,
|
28 |
"end": segment.end,
|
29 |
"text": segment.text
|
30 |
})
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
print(f"Error writing transcription file: {e}")
|
40 |
-
raise
|
41 |
-
return output_file
|
42 |
-
except FileNotFoundError as e:
|
43 |
-
print(f"Error: {e}")
|
44 |
raise
|
45 |
-
|
46 |
-
print(f"An unexpected error occurred: {e}")
|
47 |
-
raise
|
|
|
2 |
from faster_whisper import WhisperModel
|
3 |
|
4 |
class TranscriptionMaker():
|
5 |
+
#書き起こしファイルを吐き出すディレクトリを指定
|
6 |
def __init__(self,output_dir=os.path.abspath("/tmp/data/transcriptions")):
|
7 |
self.model = WhisperModel("base", device="cpu")
|
8 |
self.output_dir = output_dir
|
|
|
13 |
print(f"Error creating directory {self.output_dir}: {e}")
|
14 |
raise
|
15 |
|
16 |
+
#音声ファイルのディレクトリを受け取り、書き起こしファイルを作成する
|
17 |
+
def create_transcription(self,audio_directory):
|
18 |
+
results = []
|
19 |
+
#ディレクトリ内のファイルを全て取得
|
20 |
+
if not os.path.isdir(audio_directory):
|
21 |
+
raise ValueError(f"The specified path is not a valid directory: {audio_directory}")
|
22 |
+
audio_files = os.listdir(audio_directory)
|
23 |
+
for audio_file in audio_files:
|
24 |
+
if os.path.splitext(audio_file)[-1].lower() != '.wav':
|
25 |
+
continue
|
26 |
+
audio_path = os.path.join(audio_directory, audio_file)
|
27 |
+
try:
|
28 |
+
segments,info = list(self.model.transcribe(audio_path))
|
29 |
+
except Exception as e:
|
30 |
+
print(f"Error transcripting file {audio_path}: {e}")
|
31 |
+
raise
|
32 |
for segment in segments:
|
33 |
results.append({
|
34 |
"start": segment.start,
|
35 |
"end": segment.end,
|
36 |
"text": segment.text
|
37 |
})
|
38 |
+
#ファイルの書き込み。ファイル名は"読み込みディレクトリ名_transcription.txt"
|
39 |
+
output_file=os.path.join(self.output_dir,os.path.basename(audio_directory)+"_transcription.txt")
|
40 |
+
try:
|
41 |
+
with open(output_file,"w",encoding="utf-8") as f:
|
42 |
+
for result in results:
|
43 |
+
f.write(f"{result['text']}\n")
|
44 |
+
except OSError as e:
|
45 |
+
print(f"Error writing transcription file: {e}")
|
|
|
|
|
|
|
|
|
|
|
46 |
raise
|
47 |
+
return output_file
|
|
|
|