Tech blog
Easily synthesize subtitles into videos! Subtitle workflow created with speech recognition API
Published: 2026.01.30 Last updated: 2026.01.30
Spice Dog
Hello, I'm Spice Dog.
When did the word "blog" first come into being? It replaced the endlessly complicated method of "running a website" that had existed until then, and became a revolutionary way of disseminating information that required neither a domain nor HTML knowledge. Since then, it has taken on a more glamorous approach, centered around social media, and incorporating photos, and now we have "VLogs," where people publish their personal diaries and records in video format. It feels like digital communication is becoming closer to a human interface.
It is now common for videos, including VLogs, to have subtitles and subtitles as a means of communication. Many users broadcast globally, and many have subtitles in multiple languages. It is amazing how accessibility has improved in recent years.
Creating a video with subtitles is actually very easy using AmiVoice. Today, I'll show you an example of how to create subtitles using AmiVoice.
Structure of subtitled videos
MP4, MOV, AVI, or MPEG: which video formats have you seen? For me, it's overwhelmingly MPEG. Back when I was designing optical disc drives and video cameras, all DVD-Video formats were MPEG or VOB. While MPEG has become rare in this day and age of smartphone recording and online streaming, terrestrial digital broadcasting still uses the MPEG-2 codec, making it a long-lasting format.
However, the formats that are most suitable for subtitle generation are container types such as MP4 and MOV. Unlike MPEG, MP4 and MOV already include metadata fields for “subtitles” and also provide flexible timestamp handling. From here, the explanation will focus on methods for embedding subtitles using MP4 as the base format.
There are two methods of subtitle synthesis: hard sub (burned-in) and softsub (embedded subtitle track). The differences between them are roughly as follows:
- Hard Subtitles (hard subs / burned‑in subtitles):
- The subtitles are burned directly into the video, so viewers cannot turn them on or off.
- It looks the same on all devices and players, so there are no compatibility issues.
- Once burned in, it becomes difficult to edit the subtitles.
- Soft Subtitles (soft subs / subtitle tracks embedded in the container):
- Subtitles are embedded in the video file as a separate track and can be turned on or off by the viewer.
- By editing the subtitle file, you can easily change the content or add multiple languages.
- Soft subs may not display correctly on some devices or players.
In this global smartphone era, I feel that soft subs are more flexible and better. The only difference in how you create subtitled videos is the option settings. This time, I'll go with soft subs.
procedure
There are four steps (including watching)
STEP 1: Creating an audio file: Extracting audio from a video
STEP 2: Text Conversion: Convert audio into text using voice recognition
STEP 3: Creating subtitles and compositing videos: Embedding subtitle text into the video track
FINAL: Watch the video: You can read the subtitles
STEP 1 Creating an audio file: Extracting audio from video
The subject will be Prime Minister Takaichi's telephone conversation with US President Trump.
https://www.kantei.go.jp/jp/104/statement/2025/111125kaiken.html
Although it is possible to play a video on the web and pick up the audio directly, it is much faster to download the video and extract the audio (export it to an audio file). To do this, we use the free tool ffmpeg.
Example command to save as mp3
ffmpeg -i “%SelectedFile%” -vn -ar 44100 -ac 2 -b:a 192k “%SoundFilePath%” -y
ffmpeg parameter explanation
-i "%MovieFilePath%"
入力する動画ファイルを指定します。（例：C:\video\source.mp4）です。
-vn
Video None の略で、動画ストリームを無効化します。→ 音声だけを抽出・出力したいときに使います。
-ar 44100 Audio Rate（サンプリングレート）を 44,100 Hz に設定します。→ CD音質の標準的なサンプリングレートです。
-ac 2
Audio Channels を 2（ステレオ） に設定します。→ モノラルにしたい場合は -ac 1。
-b:a 192k Audio bitrate を 192 kbps に設定します。→ 128k～192k は音楽でも比較的良好な品質です。
"%SoundFilePath%"
出力音声ファイルのパスを指定します。拡張子（例：.mp3, .aac, .wav）によって音声コーデックやコンテナが自動選択されます。
-y 既存の出力ファイルがある場合でも確認なしで上書きします。→ 安全にしたい場合は -n（上書き禁止）に変更。
Save the input (downloaded) video “%selectedfile%” as “takaichi_voice.mp4” and the output audio file “%soundfilepath%” as “takaichi_voice.mp3”.
Tip: For audio files posted on the web as links to mp3 files, there is also a one-stop method for voice recognition and download (STEP 1 and STEP 2). We hope to release this method soon.
STEP 2: Convert to text: Convert to text using voice recognition
We will convert speech to text using the AmiVoice API from the AmiVoice Cloud Platform (ACP). It's easy to learn how to use the API by using the sample program in the ACP manual. Here, we'll use WrpSimpleTester.py.
How to obtain | AmiVoice Cloud Platform
WrpSimpleTester.py
Lines 105 and 106 are commented out to extract only JSON.
# encoding: UTF-8
import sys
import time
import json
# <!-- バイトコードキャッシュファイルの作成を抑制するために...
sys.dont_write_bytecode = True
# -->
import com.amivoice.wrp.Wrp
import com.amivoice.wrp.WrpListener
class WrpSimpleTester(com.amivoice.wrp.WrpListener):
@staticmethod
def main(args):
if len(args) < 4:
print("Usage: python WrpSimpleTester.py <url> <audioFileName> <codec> <grammarFileNames> [<authorization>]")
return
serverURL = args[0]
audioFileName = args[1]
codec = args[2]
grammarFileNames = args[3]
authorization = args[4] if len(args) > 4 else None
# WebSocket 音声認識サーバイベントリスナの作成
listener = WrpSimpleTester()
# WebSocket 音声認識サーバの初期化
wrp = com.amivoice.wrp.Wrp.construct()
wrp.setListener(listener)
wrp.setServerURL(serverURL)
wrp.setCodec(codec)
wrp.setGrammarFileNames(grammarFileNames)
wrp.setAuthorization(authorization)
# WebSocket 音声認識サーバへの接続
if not wrp.connect():
print(wrp.getLastMessage())
print(u"WebSocket 音声認識サーバ %s への接続に失敗しました。" % serverURL)
return
try:
# WebSocket 音声認識サーバへの音声データの送信の開始
if not wrp.feedDataResume():
print(wrp.getLastMessage())
print(u"WebSocket 音声認識サーバへの音声データの送信の開始に失敗しました。")
return
try:
with open(audioFileName, "rb") as audioStream:
# 音声データファイルからの音声データの読み込み
audioData = audioStream.read(4096)
while len(audioData) > 0:
# 微小時間のスリープ
wrp.sleep(1)
# 認識結果情報待機数が 1 以下になるまでスリープ
maxSleepTime = 50000
while wrp.getWaitingResults() > 1 and maxSleepTime > 0:
wrp.sleep(100)
maxSleepTime -= 100
# WebSocket 音声認識サーバへの音声データの送信
if not wrp.feedData(audioData, 0, len(audioData)):
print(wrp.getLastMessage())
print(u"WebSocket 音声認識サーバへの音声データの送信に失敗しました。")
break
# 音声データファイルからの音声データの読み込み
audioData = audioStream.read(4096)
except:
print(u"音声データファイル %s の読み込みに失敗しました。" % audioFileName)
# WebSocket 音声認識サーバへの音声データの送信の完了
if not wrp.feedDataPause():
print(wrp.getLastMessage())
print(u"WebSocket 音声認識サーバへの音声データの送信の完了に失敗しました。")
return
finally:
# WebSocket 音声認識サーバからの切断
wrp.disconnect()
def __init__(self):
pass
def utteranceStarted(self, startTime):
# print("S %d" % startTime)
pass
def utteranceEnded(self, endTime):
# print("E %d" % endTime)
pass
def resultCreated(self):
# print("C")
pass
def resultUpdated(self, result):
# print("U %s" % result)
pass
def resultFinalized(self, result):
# print("F %s" % result)
print(result)
text = self.text_(result)
#if text != None: # コメントアウトしているのは、JSON以外の音声認識結果を表示しないため
# print(" -> %s" % text)
def eventNotified(self, eventId, eventMessage):
# print(eventId + " " + eventMessage)
pass
def TRACE(self, message):
pass
def text_(self, result):
try:
return json.loads(result)["text"]
except:
return None
WrpSimpleTester.main(sys.argv[1:])
There are two methods for making speech recognition requests: "synchronous" and "asynchronous." You can use either interface, but for audio files exceeding 16MB, please use "asynchronous." Please note that all recognition results will be obtained in JSON format, including a timestamp.
Here is an example of how to run it in a batch file for Windows OS (using a 16k general-purpose engine):
run_amivoice.bat
Argument %1: Audio file path
Argument %2: JSON file path to output
set PYTHONPATH=src
set SSL_CERT_FILE=../../curl-ca-bundle.crt
python WrpSimpleTesterJimaku.py wss://acp-api.amivoice.com/v1/ %1 16K -a-general {APPKEY} |jq > %2
@pauseThe conversation text obtained in JSON format is here
The file name should be “takaichi_voice.json”.
takaichi_voice.json
{
"results": [
{
"tokens": [
{
"written": "皆様",
"confidence": 0.80,
"starttime": 422,
"endtime": 694,
"spoken": "みなさま"
},
{
"written": "どうぞ",
"confidence": 1.00,
"starttime": 694,
"endtime": 1286,
"spoken": "どうぞ"
}
],
"confidence": 0.922,
"starttime": 150,
"endtime": 1286,
"tags": [],
"rulename": "",
"text": "皆様どうぞ"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183844",
"text": "皆様どうぞ",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "朝日新聞",
"confidence": 0.98,
"starttime": 1938,
"endtime": 2562,
"spoken": "あさひしんぶん"
},
{
"written": "陣野",
"confidence": 0.98,
"starttime": 2562,
"endtime": 2786,
"spoken": "じんの"
},
{
"written": "と",
"confidence": 1.00,
"starttime": 2786,
"endtime": 2882,
"spoken": "と"
},
{
"written": "申し",
"confidence": 1.00,
"starttime": 2882,
"endtime": 3106,
"spoken": "もうし"
},
{
"written": "ます",
"confidence": 0.96,
"starttime": 3106,
"endtime": 3378,
"spoken": "ます"
},
{
"written": "。",
"confidence": 0.58,
"starttime": 3378,
"endtime": 3410,
"spoken": "_"
},
{
"written": "よろしくお願いします",
"confidence": 0.99,
"starttime": 3410,
"endtime": 4386,
"spoken": "よろしくおねがいします"
},
{
"written": "。",
"confidence": 0.53,
"starttime": 4386,
"endtime": 4418,
"spoken": "_"
},
{
"written": "よろしくお願いします",
"confidence": 0.98,
"starttime": 4418,
"endtime": 5522,
"spoken": "よろしくおねがいします"
},
{
"written": "。",
"confidence": 0.99,
"starttime": 5522,
"endtime": 5634,
"spoken": "_"
},
{
"written": "総理",
"confidence": 1.00,
"starttime": 5634,
"endtime": 5970,
"spoken": "そうり"
},
{
"written": "は",
"confidence": 1.00,
"starttime": 5970,
"endtime": 6130,
"spoken": "は"
},
{
"written": "アメリカ",
"confidence": 1.00,
"starttime": 6130,
"endtime": 6498,
"spoken": "あめりか"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 6498,
"endtime": 6610,
"spoken": "の"
},
{
"written": "トランプ",
"confidence": 0.62,
"starttime": 6610,
"endtime": 6946,
"spoken": "とらんぷ"
},
{
"written": "大統領",
"confidence": 0.62,
"starttime": 6946,
"endtime": 7474,
"spoken": "だいとうりょう"
},
{
"written": "と",
"confidence": 1.00,
"starttime": 7474,
"endtime": 7618,
"spoken": "と"
},
{
"written": "電話",
"confidence": 1.00,
"starttime": 7650,
"endtime": 7954,
"spoken": "でんわ"
},
{
"written": "会談",
"confidence": 1.00,
"starttime": 7954,
"endtime": 8290,
"spoken": "かいだん"
},
{
"written": "に",
"confidence": 0.96,
"starttime": 8290,
"endtime": 8370,
"spoken": "に"
},
{
"written": "臨ま",
"confidence": 0.70,
"starttime": 8402,
"endtime": 8738,
"spoken": "のぞま"
},
{
"written": "れる",
"confidence": 1.00,
"starttime": 8738,
"endtime": 8930,
"spoken": "れる"
},
{
"written": "と",
"confidence": 1.00,
"starttime": 8930,
"endtime": 9010,
"spoken": "と"
},
{
"written": "行動",
"confidence": 0.99,
"starttime": 9010,
"endtime": 9346,
"spoken": "こうどう"
},
{
"written": "が",
"confidence": 1.00,
"starttime": 9346,
"endtime": 9458,
"spoken": "が"
},
{
"written": "出て",
"confidence": 0.95,
"starttime": 9458,
"endtime": 9634,
"spoken": "でて"
},
{
"written": "います",
"confidence": 0.98,
"starttime": 9634,
"endtime": 10194,
"spoken": "います"
},
{
"written": "。",
"confidence": 0.98,
"starttime": 10194,
"endtime": 10290,
"spoken": "_"
}
],
"confidence": 0.978,
"starttime": 1650,
"endtime": 10290,
"tags": [],
"rulename": "",
"text": "朝日新聞陣野と申します。よろしくお願いします。よろしくお願いします。総理はアメリカのトランプ大統領と電話会談に臨まれると行動が出ています。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183845",
"text": "朝日新聞陣野と申します。よろしくお願いします。よろしくお願いします。総理はアメリカのトランプ大統領と電話会談に臨まれると行動が出ています。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "こちら",
"confidence": 1.00,
"starttime": 10332,
"endtime": 10620,
"spoken": "こちら"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 10620,
"endtime": 10732,
"spoken": "の"
},
{
"written": "実施",
"confidence": 1.00,
"starttime": 10732,
"endtime": 11084,
"spoken": "じっし"
},
{
"written": "状況",
"confidence": 1.00,
"starttime": 11084,
"endtime": 11436,
"spoken": "じょうきょう"
},
{
"written": "と",
"confidence": 1.00,
"starttime": 11436,
"endtime": 11708,
"spoken": "と"
},
{
"written": "どちら",
"confidence": 1.00,
"starttime": 11740,
"endtime": 12108,
"spoken": "どちら"
},
{
"written": "から",
"confidence": 1.00,
"starttime": 12108,
"endtime": 12252,
"spoken": "から"
},
{
"written": "持ちかけ",
"confidence": 1.00,
"starttime": 12252,
"endtime": 12652,
"spoken": "もちかけ"
},
{
"written": "られ",
"confidence": 1.00,
"starttime": 12652,
"endtime": 12844,
"spoken": "られ"
},
{
"written": "た",
"confidence": 1.00,
"starttime": 12844,
"endtime": 12972,
"spoken": "た"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 12972,
"endtime": 13052,
"spoken": "の"
},
{
"written": "か",
"confidence": 1.00,
"starttime": 13052,
"endtime": 13164,
"spoken": "か"
},
{
"written": "について",
"confidence": 1.00,
"starttime": 13164,
"endtime": 13676,
"spoken": "について"
},
{
"written": "お尋ね",
"confidence": 0.82,
"starttime": 13708,
"endtime": 14092,
"spoken": "おたずね"
},
{
"written": "します",
"confidence": 0.97,
"starttime": 14092,
"endtime": 14764,
"spoken": "します"
},
{
"written": "。",
"confidence": 0.99,
"starttime": 14764,
"endtime": 14908,
"spoken": "_"
},
{
"written": "また",
"confidence": 1.00,
"starttime": 14908,
"endtime": 15340,
"spoken": "また"
},
{
"written": "、",
"confidence": 0.80,
"starttime": 15340,
"endtime": 15372,
"spoken": "_"
},
{
"written": "既に",
"confidence": 1.00,
"starttime": 15372,
"endtime": 15724,
"spoken": "すでに"
},
{
"written": "実施",
"confidence": 1.00,
"starttime": 15724,
"endtime": 16012,
"spoken": "じっし"
},
{
"written": "済み",
"confidence": 1.00,
"starttime": 16012,
"endtime": 16204,
"spoken": "ずみ"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 16204,
"endtime": 16300,
"spoken": "の"
},
{
"written": "場合",
"confidence": 1.00,
"starttime": 16300,
"endtime": 16556,
"spoken": "ばあい"
},
{
"written": "は",
"confidence": 1.00,
"starttime": 16588,
"endtime": 16812,
"spoken": "は"
},
{
"written": "、",
"confidence": 0.52,
"starttime": 16812,
"endtime": 16844,
"spoken": "_"
},
{
"written": "会談",
"confidence": 0.80,
"starttime": 16844,
"endtime": 17196,
"spoken": "かいだん"
},
{
"written": "で",
"confidence": 1.00,
"starttime": 17196,
"endtime": 17292,
"spoken": "で"
},
{
"written": "どの",
"confidence": 0.99,
"starttime": 17292,
"endtime": 17484,
"spoken": "どの"
},
{
"written": "よう",
"confidence": 1.00,
"starttime": 17484,
"endtime": 17596,
"spoken": "よう"
},
{
"written": "な",
"confidence": 1.00,
"starttime": 17596,
"endtime": 17676,
"spoken": "な"
},
{
"written": "議論",
"confidence": 1.00,
"starttime": 17676,
"endtime": 17980,
"spoken": "ぎろん"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 17980,
"endtime": 18076,
"spoken": "を"
},
{
"written": "され",
"confidence": 1.00,
"starttime": 18076,
"endtime": 18252,
"spoken": "され"
},
{
"written": "た",
"confidence": 1.00,
"starttime": 18252,
"endtime": 18380,
"spoken": "た"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 18380,
"endtime": 18492,
"spoken": "の"
},
{
"written": "でしょうか？",
"confidence": 1.00,
"starttime": 18492,
"endtime": 19132,
"spoken": "でしょうか_"
}
],
"confidence": 0.972,
"starttime": 10300,
"endtime": 19132,
"tags": [],
"rulename": "",
"text": "こちらの実施状況とどちらから持ちかけられたのかについてお尋ねします。また、既に実施済みの場合は、会談でどのような議論をされたのでしょうか？"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183846",
"text": "こちらの実施状況とどちらから持ちかけられたのかについてお尋ねします。また、既に実施済みの場合は、会談でどのような議論をされたのでしょうか？",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "台湾",
"confidence": 1.00,
"starttime": 19326,
"endtime": 19790,
"spoken": "たいわん"
},
{
"written": "問題",
"confidence": 1.00,
"starttime": 19790,
"endtime": 20206,
"spoken": "もんだい"
},
{
"written": "や",
"confidence": 1.00,
"starttime": 20206,
"endtime": 20478,
"spoken": "や"
},
{
"written": "総理",
"confidence": 1.00,
"starttime": 20510,
"endtime": 20830,
"spoken": "そうり"
},
{
"written": "の",
"confidence": 0.87,
"starttime": 20830,
"endtime": 20958,
"spoken": "の"
},
{
"written": "存立",
"confidence": 0.97,
"starttime": 21038,
"endtime": 21486,
"spoken": "そんりつ"
},
{
"written": "危機",
"confidence": 1.00,
"starttime": 21486,
"endtime": 21710,
"spoken": "きき"
},
{
"written": "事態",
"confidence": 1.00,
"starttime": 21710,
"endtime": 22078,
"spoken": "じたい"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 22078,
"endtime": 22110,
"spoken": "を"
},
{
"written": "巡る",
"confidence": 1.00,
"starttime": 22110,
"endtime": 22382,
"spoken": "めぐる"
},
{
"written": "答弁",
"confidence": 1.00,
"starttime": 22382,
"endtime": 22750,
"spoken": "とうべん"
},
{
"written": "について",
"confidence": 1.00,
"starttime": 22750,
"endtime": 23358,
"spoken": "について"
},
{
"written": "、",
"confidence": 0.56,
"starttime": 23358,
"endtime": 23390,
"spoken": "_"
},
{
"written": "どの",
"confidence": 1.00,
"starttime": 23390,
"endtime": 23582,
"spoken": "どの"
},
{
"written": "よう",
"confidence": 1.00,
"starttime": 23582,
"endtime": 23710,
"spoken": "よう"
},
{
"written": "な",
"confidence": 1.00,
"starttime": 23710,
"endtime": 23806,
"spoken": "な"
},
{
"written": "やり取り",
"confidence": 1.00,
"starttime": 23806,
"endtime": 24254,
"spoken": "やりとり"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 24254,
"endtime": 24350,
"spoken": "を"
},
{
"written": "され",
"confidence": 1.00,
"starttime": 24350,
"endtime": 24574,
"spoken": "され"
},
{
"written": "た",
"confidence": 1.00,
"starttime": 24574,
"endtime": 24686,
"spoken": "た"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 24686,
"endtime": 24798,
"spoken": "の"
},
{
"written": "か",
"confidence": 1.00,
"starttime": 24798,
"endtime": 24910,
"spoken": "か"
},
{
"written": "について",
"confidence": 0.99,
"starttime": 24910,
"endtime": 25278,
"spoken": "について"
},
{
"written": "も",
"confidence": 1.00,
"starttime": 25278,
"endtime": 25470,
"spoken": "も"
},
{
"written": "あわせ",
"confidence": 0.77,
"starttime": 25470,
"endtime": 25774,
"spoken": "あわせ"
},
{
"written": "て",
"confidence": 1.00,
"starttime": 25774,
"endtime": 25870,
"spoken": "て"
},
{
"written": "伺い",
"confidence": 1.00,
"starttime": 25870,
"endtime": 26206,
"spoken": "うかがい"
},
{
"written": "ます",
"confidence": 1.00,
"starttime": 26206,
"endtime": 26766,
"spoken": "ます"
},
{
"written": "。",
"confidence": 0.98,
"starttime": 26766,
"endtime": 26846,
"spoken": "_"
}
],
"confidence": 0.987,
"starttime": 19150,
"endtime": 26846,
"tags": [],
"rulename": "",
"text": "台湾問題や総理の存立危機事態を巡る答弁について、どのようなやり取りをされたのかについてもあわせて伺います。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183848",
"text": "台湾問題や総理の存立危機事態を巡る答弁について、どのようなやり取りをされたのかについてもあわせて伺います。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "はい",
"confidence": 1.00,
"starttime": 27338,
"endtime": 27802,
"spoken": "はい"
},
{
"written": "。",
"confidence": 0.77,
"starttime": 27802,
"endtime": 27994,
"spoken": "_"
}
],
"confidence": 1.000,
"starttime": 27050,
"endtime": 27994,
"tags": [],
"rulename": "",
"text": "はい。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183850",
"text": "はい。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "ありがとうございます",
"confidence": 1.00,
"starttime": 28112,
"endtime": 29152,
"spoken": "ありがとうございます"
},
{
"written": "。",
"confidence": 0.84,
"starttime": 29312,
"endtime": 29376,
"spoken": "_"
},
{
"written": "先ほど",
"confidence": 1.00,
"starttime": 29664,
"endtime": 30448,
"spoken": "さきほど"
},
{
"written": "トランプ大統領",
"confidence": 1.00,
"starttime": 30448,
"endtime": 31424,
"spoken": "とらんぷだいとうりょう"
},
{
"written": "から",
"confidence": 1.00,
"starttime": 31424,
"endtime": 31744,
"spoken": "から"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 31744,
"endtime": 32048,
"spoken": "の"
},
{
"written": "お",
"confidence": 1.00,
"starttime": 32304,
"endtime": 32416,
"spoken": "お"
},
{
"written": "申し出",
"confidence": 1.00,
"starttime": 32416,
"endtime": 32896,
"spoken": "もうしで"
},
{
"written": "に",
"confidence": 1.00,
"starttime": 32896,
"endtime": 33008,
"spoken": "に"
},
{
"written": "より",
"confidence": 1.00,
"starttime": 33008,
"endtime": 33200,
"spoken": "より"
},
{
"written": "まして",
"confidence": 1.00,
"starttime": 33200,
"endtime": 33600,
"spoken": "まして"
},
{
"written": "電話",
"confidence": 0.99,
"starttime": 33600,
"endtime": 33936,
"spoken": "でんわ"
},
{
"written": "会談",
"confidence": 1.00,
"starttime": 33936,
"endtime": 34448,
"spoken": "かいだん"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 34448,
"endtime": 34608,
"spoken": "を"
},
{
"written": "行い",
"confidence": 1.00,
"starttime": 34608,
"endtime": 35040,
"spoken": "おこない"
},
{
"written": "ました",
"confidence": 1.00,
"starttime": 35040,
"endtime": 35440,
"spoken": "ました"
},
{
"written": "。",
"confidence": 0.96,
"starttime": 35440,
"endtime": 35584,
"spoken": "_"
}
],
"confidence": 0.998,
"starttime": 28000,
"endtime": 35584,
"tags": [],
"rulename": "",
"text": "ありがとうございます。先ほどトランプ大統領からのお申し出によりまして電話会談を行いました。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183850_",
"text": "ありがとうございます。先ほどトランプ大統領からのお申し出によりまして電話会談を行いました。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "私",
"confidence": 1.00,
"starttime": 36144,
"endtime": 36528,
"spoken": "わたくし"
},
{
"written": "から",
"confidence": 1.00,
"starttime": 36528,
"endtime": 36720,
"spoken": "から"
},
{
"written": "は",
"confidence": 1.00,
"starttime": 36720,
"endtime": 37088,
"spoken": "は"
},
{
"written": "トランプ大統領",
"confidence": 1.00,
"starttime": 37376,
"endtime": 38336,
"spoken": "とらんぷだいとうりょう"
},
{
"written": "が",
"confidence": 1.00,
"starttime": 38336,
"endtime": 38656,
"spoken": "が"
},
{
"written": "訪日",
"confidence": 1.00,
"starttime": 38864,
"endtime": 39424,
"spoken": "ほうにち"
},
{
"written": "され",
"confidence": 1.00,
"starttime": 39424,
"endtime": 39680,
"spoken": "され"
},
{
"written": "た",
"confidence": 1.00,
"starttime": 39680,
"endtime": 40000,
"spoken": "た"
},
{
"written": "とき",
"confidence": 0.81,
"starttime": 40224,
"endtime": 40496,
"spoken": "とき"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 40496,
"endtime": 40592,
"spoken": "の"
},
{
"written": "こと",
"confidence": 0.89,
"starttime": 40592,
"endtime": 40848,
"spoken": "こと"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 40848,
"endtime": 40944,
"spoken": "を"
},
{
"written": "報道",
"confidence": 1.00,
"starttime": 40944,
"endtime": 41344,
"spoken": "ほうどう"
},
{
"written": "した",
"confidence": 1.00,
"starttime": 41344,
"endtime": 41744,
"spoken": "した"
},
{
"written": "。",
"confidence": 0.92,
"starttime": 41744,
"endtime": 41936,
"spoken": "_"
}
],
"confidence": 0.995,
"starttime": 35600,
"endtime": 41936,
"tags": [],
"rulename": "",
"text": "私からはトランプ大統領が訪日されたときのことを報道した。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183851",
"text": "私からはトランプ大統領が訪日されたときのことを報道した。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "アメリカ",
"confidence": 1.00,
"starttime": 42526,
"endtime": 43022,
"spoken": "あめりか"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 43022,
"endtime": 43374,
"spoken": "の"
},
{
"written": "新聞",
"confidence": 1.00,
"starttime": 43662,
"endtime": 44094,
"spoken": "しんぶん"
},
{
"written": "に",
"confidence": 1.00,
"starttime": 44094,
"endtime": 44430,
"spoken": "に"
},
{
"written": "採用",
"confidence": 0.71,
"starttime": 44558,
"endtime": 45134,
"spoken": "さいよう"
},
{
"written": "した",
"confidence": 0.99,
"starttime": 45134,
"endtime": 45374,
"spoken": "した"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 45374,
"endtime": 45502,
"spoken": "の"
},
{
"written": "を",
"confidence": 0.48,
"starttime": 45534,
"endtime": 45566,
"spoken": "を"
},
{
"written": "送っ",
"confidence": 1.00,
"starttime": 46110,
"endtime": 46430,
"spoken": "おくっ"
},
{
"written": "て",
"confidence": 1.00,
"starttime": 46430,
"endtime": 46526,
"spoken": "て"
},
{
"written": "きて",
"confidence": 1.00,
"starttime": 46526,
"endtime": 46702,
"spoken": "きて"
},
{
"written": "くださっ",
"confidence": 1.00,
"starttime": 46702,
"endtime": 47166,
"spoken": "くださっ"
},
{
"written": "た",
"confidence": 1.00,
"starttime": 47166,
"endtime": 47262,
"spoken": "た"
},
{
"written": "こと",
"confidence": 0.76,
"starttime": 47262,
"endtime": 47518,
"spoken": "こと"
},
{
"written": "への",
"confidence": 1.00,
"starttime": 47518,
"endtime": 47758,
"spoken": "への"
},
{
"written": "お礼",
"confidence": 0.57,
"starttime": 47758,
"endtime": 48046,
"spoken": "おれい"
},
{
"written": "と",
"confidence": 1.00,
"starttime": 48046,
"endtime": 48366,
"spoken": "と"
}
],
"confidence": 0.983,
"starttime": 41950,
"endtime": 48542,
"tags": [],
"rulename": "",
"text": "アメリカの新聞に採用したのを送ってきてくださったことへのお礼と"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183852",
"text": "アメリカの新聞に採用したのを送ってきてくださったことへのお礼と",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "あと",
"confidence": 0.78,
"starttime": 48582,
"endtime": 48838,
"spoken": "あと"
},
{
"written": "ウクライナ",
"confidence": 0.82,
"starttime": 49030,
"endtime": 49638,
"spoken": "うくらいな"
},
{
"written": "和平",
"confidence": 0.99,
"starttime": 49638,
"endtime": 50182,
"spoken": "わへい"
},
{
"written": "に",
"confidence": 1.00,
"starttime": 50214,
"endtime": 50422,
"spoken": "に"
},
{
"written": "関する",
"confidence": 1.00,
"starttime": 50422,
"endtime": 51238,
"spoken": "かんする"
},
{
"written": "米国",
"confidence": 1.00,
"starttime": 51782,
"endtime": 52294,
"spoken": "べいこく"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 52294,
"endtime": 52486,
"spoken": "の"
},
{
"written": "取り組み",
"confidence": 1.00,
"starttime": 52518,
"endtime": 53078,
"spoken": "とりくみ"
},
{
"written": "への",
"confidence": 1.00,
"starttime": 53078,
"endtime": 53302,
"spoken": "への"
},
{
"written": "評価",
"confidence": 1.00,
"starttime": 53302,
"endtime": 53846,
"spoken": "ひょうか"
}
],
"confidence": 0.991,
"starttime": 48550,
"endtime": 54086,
"tags": [],
"rulename": "",
"text": "あとウクライナ和平に関する米国の取り組みへの評価"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183853",
"text": "あとウクライナ和平に関する米国の取り組みへの評価",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "について",
"confidence": 1.00,
"starttime": 54276,
"endtime": 55252,
"spoken": "について"
},
{
"written": "お伝え",
"confidence": 1.00,
"starttime": 55604,
"endtime": 56084,
"spoken": "おつたえ"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 56084,
"endtime": 56164,
"spoken": "を"
},
{
"written": "しました",
"confidence": 1.00,
"starttime": 56164,
"endtime": 56740,
"spoken": "しました"
},
{
"written": "その",
"confidence": 1.00,
"starttime": 57108,
"endtime": 57380,
"spoken": "その"
},
{
"written": "上",
"confidence": 1.00,
"starttime": 57380,
"endtime": 57524,
"spoken": "うえ"
},
{
"written": "で",
"confidence": 1.00,
"starttime": 57524,
"endtime": 57716,
"spoken": "で"
},
{
"written": "日米",
"confidence": 1.00,
"starttime": 57716,
"endtime": 58196,
"spoken": "にちべい"
},
{
"written": "同盟",
"confidence": 1.00,
"starttime": 58196,
"endtime": 58484,
"spoken": "どうめい"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 58484,
"endtime": 58580,
"spoken": "の"
},
{
"written": "強化",
"confidence": 1.00,
"starttime": 58580,
"endtime": 58996,
"spoken": "きょうか"
},
{
"written": "や",
"confidence": 1.00,
"starttime": 58996,
"endtime": 59300,
"spoken": "や"
},
{
"written": "、",
"confidence": 0.75,
"starttime": 59412,
"endtime": 59492,
"spoken": "_"
}
],
"confidence": 0.999,
"starttime": 54100,
"endtime": 59492,
"tags": [],
"rulename": "",
"text": "についてお伝えをしましたその上で日米同盟の強化や、"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183854",
"text": "についてお伝えをしましたその上で日米同盟の強化や、",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "インド",
"confidence": 0.94,
"starttime": 59612,
"endtime": 59916,
"spoken": "いんど"
},
{
"written": "太平洋",
"confidence": 0.94,
"starttime": 59916,
"endtime": 60380,
"spoken": "たいへいよう"
},
{
"written": "地域",
"confidence": 0.94,
"starttime": 60380,
"endtime": 60732,
"spoken": "ちいき"
},
{
"written": "が",
"confidence": 1.00,
"starttime": 60732,
"endtime": 60828,
"spoken": "が"
},
{
"written": "直面",
"confidence": 1.00,
"starttime": 60828,
"endtime": 61324,
"spoken": "ちょくめん"
},
{
"written": "する",
"confidence": 1.00,
"starttime": 61324,
"endtime": 61692,
"spoken": "する"
},
{
"written": "情勢",
"confidence": 1.00,
"starttime": 61692,
"endtime": 62108,
"spoken": "じょうせい"
},
{
"written": "や",
"confidence": 1.00,
"starttime": 62108,
"endtime": 62396,
"spoken": "や"
},
{
"written": "諸",
"confidence": 1.00,
"starttime": 62428,
"endtime": 62556,
"spoken": "しょ"
},
{
"written": "課題",
"confidence": 1.00,
"starttime": 62556,
"endtime": 62876,
"spoken": "かだい"
},
{
"written": "について",
"confidence": 1.00,
"starttime": 62876,
"endtime": 63372,
"spoken": "について"
},
{
"written": "幅広く",
"confidence": 1.00,
"starttime": 63372,
"endtime": 64092,
"spoken": "はばひろく"
},
{
"written": "意見",
"confidence": 1.00,
"starttime": 64380,
"endtime": 64732,
"spoken": "いけん"
},
{
"written": "交換",
"confidence": 1.00,
"starttime": 64732,
"endtime": 65212,
"spoken": "こうかん"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 65212,
"endtime": 65468,
"spoken": "を"
},
{
"written": "行い",
"confidence": 1.00,
"starttime": 65500,
"endtime": 65868,
"spoken": "おこない"
},
{
"written": "ました",
"confidence": 1.00,
"starttime": 65868,
"endtime": 66236,
"spoken": "ました"
},
{
"written": "。",
"confidence": 0.91,
"starttime": 66236,
"endtime": 66444,
"spoken": "_"
}
],
"confidence": 0.999,
"starttime": 59500,
"endtime": 66444,
"tags": [],
"rulename": "",
"text": "インド太平洋地域が直面する情勢や諸課題について幅広く意見交換を行いました。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183855",
"text": "インド太平洋地域が直面する情勢や諸課題について幅広く意見交換を行いました。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "その",
"confidence": 1.00,
"starttime": 66738,
"endtime": 66962,
"spoken": "その"
},
{
"written": "中",
"confidence": 0.88,
"starttime": 66962,
"endtime": 67218,
"spoken": "なか"
},
{
"written": "で",
"confidence": 1.00,
"starttime": 67218,
"endtime": 67442,
"spoken": "で"
},
{
"written": "、",
"confidence": 0.63,
"starttime": 67442,
"endtime": 67474,
"spoken": "_"
},
{
"written": "トランプ大統領",
"confidence": 0.97,
"starttime": 67474,
"endtime": 68354,
"spoken": "とらんぷだいとうりょう"
},
{
"written": "から",
"confidence": 0.99,
"starttime": 68354,
"endtime": 68626,
"spoken": "から"
},
{
"written": "は",
"confidence": 1.00,
"starttime": 68626,
"endtime": 68818,
"spoken": "は"
},
{
"written": "昨晩",
"confidence": 1.00,
"starttime": 68818,
"endtime": 69458,
"spoken": "さくばん"
},
{
"written": "行わ",
"confidence": 1.00,
"starttime": 69922,
"endtime": 70482,
"spoken": "おこなわ"
},
{
"written": "れ",
"confidence": 1.00,
"starttime": 70482,
"endtime": 70610,
"spoken": "れ"
},
{
"written": "た",
"confidence": 1.00,
"starttime": 70610,
"endtime": 70930,
"spoken": "た"
},
{
"written": "米中",
"confidence": 0.99,
"starttime": 71186,
"endtime": 71586,
"spoken": "べいちゅう"
},
{
"written": "首脳",
"confidence": 1.00,
"starttime": 71586,
"endtime": 71954,
"spoken": "しゅのう"
},
{
"written": "会談",
"confidence": 1.00,
"starttime": 71954,
"endtime": 72338,
"spoken": "かいだん"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 72338,
"endtime": 72434,
"spoken": "を"
},
{
"written": "含む",
"confidence": 1.00,
"starttime": 72434,
"endtime": 73106,
"spoken": "ふくむ"
}
],
"confidence": 0.996,
"starttime": 66450,
"endtime": 73138,
"tags": [],
"rulename": "",
"text": "その中で、トランプ大統領からは昨晩行われた米中首脳会談を含む"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183856",
"text": "その中で、トランプ大統領からは昨晩行われた米中首脳会談を含む",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "最近",
"confidence": 1.00,
"starttime": 73550,
"endtime": 74142,
"spoken": "さいきん"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 74142,
"endtime": 74286,
"spoken": "の"
},
{
"written": "米中",
"confidence": 1.00,
"starttime": 74286,
"endtime": 74670,
"spoken": "べいちゅう"
},
{
"written": "関係",
"confidence": 1.00,
"starttime": 74670,
"endtime": 75150,
"spoken": "かんけい"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 75150,
"endtime": 75374,
"spoken": "の"
},
{
"written": "状況",
"confidence": 1.00,
"starttime": 75406,
"endtime": 75886,
"spoken": "じょうきょう"
},
{
"written": "について",
"confidence": 1.00,
"starttime": 75886,
"endtime": 76702,
"spoken": "について"
},
{
"written": "説明",
"confidence": 1.00,
"starttime": 77166,
"endtime": 77758,
"spoken": "せつめい"
},
{
"written": "が",
"confidence": 1.00,
"starttime": 77758,
"endtime": 78222,
"spoken": "が"
},
{
"written": "ございました",
"confidence": 1.00,
"starttime": 78494,
"endtime": 79326,
"spoken": "ございました"
},
{
"written": "。",
"confidence": 0.62,
"starttime": 79326,
"endtime": 79438,
"spoken": "_"
}
],
"confidence": 1.000,
"starttime": 73150,
"endtime": 79438,
"tags": [],
"rulename": "",
"text": "最近の米中関係の状況について説明がございました。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183857",
"text": "最近の米中関係の状況について説明がございました。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "あわせて",
"confidence": 0.79,
"starttime": 79578,
"endtime": 80314,
"spoken": "あわせて"
},
{
"written": "私",
"confidence": 1.00,
"starttime": 80634,
"endtime": 81098,
"spoken": "わたくし"
},
{
"written": "が",
"confidence": 1.00,
"starttime": 81098,
"endtime": 81338,
"spoken": "が"
},
{
"written": "出席",
"confidence": 1.00,
"starttime": 81338,
"endtime": 81882,
"spoken": "しゅっせき"
},
{
"written": "した",
"confidence": 1.00,
"starttime": 81882,
"endtime": 82314,
"spoken": "した"
},
{
"written": "GTP",
"confidence": 0.72,
"starttime": 82490,
"endtime": 83402,
"spoken": "じーてぃーぴー"
},
{
"written": "の",
"confidence": 0.72,
"starttime": 84154,
"endtime": 84378,
"spoken": "の"
},
{
"written": "様子",
"confidence": 0.57,
"starttime": 84410,
"endtime": 84842,
"spoken": "ようす"
},
{
"written": "に",
"confidence": 0.57,
"starttime": 84842,
"endtime": 85306,
"spoken": "に"
},
{
"written": "つい",
"confidence": 0.87,
"starttime": 85418,
"endtime": 85706,
"spoken": "つい"
},
{
"written": "ても",
"confidence": 0.89,
"starttime": 85706,
"endtime": 86138,
"spoken": "ても"
},
{
"written": "尋ね",
"confidence": 0.95,
"starttime": 86170,
"endtime": 86634,
"spoken": "たずね"
},
{
"written": "られ",
"confidence": 1.00,
"starttime": 86634,
"endtime": 86842,
"spoken": "られ"
},
{
"written": "ました",
"confidence": 1.00,
"starttime": 86842,
"endtime": 87210,
"spoken": "ました"
},
{
"written": "ので",
"confidence": 1.00,
"starttime": 87210,
"endtime": 87594,
"spoken": "ので"
},
{
"written": "、",
"confidence": 0.76,
"starttime": 87594,
"endtime": 87690,
"spoken": "_"
}
],
"confidence": 0.996,
"starttime": 79450,
"endtime": 87690,
"tags": [],
"rulename": "",
"text": "あわせて私が出席したGTPの様子についても尋ねられましたので、"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183857_",
"text": "あわせて私が出席したGTPの様子についても尋ねられましたので、",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "お答え",
"confidence": 1.00,
"starttime": 87764,
"endtime": 88468,
"spoken": "おこたえ"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 88468,
"endtime": 88724,
"spoken": "を"
},
{
"written": "いたし",
"confidence": 1.00,
"starttime": 88756,
"endtime": 89156,
"spoken": "いたし"
},
{
"written": "ました",
"confidence": 1.00,
"starttime": 89156,
"endtime": 89652,
"spoken": "ました"
},
{
"written": "現下",
"confidence": 1.00,
"starttime": 90036,
"endtime": 90420,
"spoken": "げんか"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 90420,
"endtime": 90628,
"spoken": "の"
},
{
"written": "国際",
"confidence": 1.00,
"starttime": 90660,
"endtime": 91060,
"spoken": "こくさい"
},
{
"written": "情勢",
"confidence": 0.98,
"starttime": 91060,
"endtime": 91380,
"spoken": "じょうせい"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 91380,
"endtime": 91540,
"spoken": "の"
},
{
"written": "もと",
"confidence": 1.00,
"starttime": 91540,
"endtime": 91860,
"spoken": "もと"
},
{
"written": "で",
"confidence": 1.00,
"starttime": 91860,
"endtime": 92244,
"spoken": "で"
},
{
"written": "先般",
"confidence": 1.00,
"starttime": 92580,
"endtime": 93092,
"spoken": "せんぱん"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 93092,
"endtime": 93332,
"spoken": "の"
},
{
"written": "トランプ大統領",
"confidence": 0.94,
"starttime": 93380,
"endtime": 94164,
"spoken": "とらんぷだいとうりょう"
},
{
"written": "委員",
"confidence": 0.98,
"starttime": 94164,
"endtime": 94292,
"spoken": "いいん"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 94292,
"endtime": 94500,
"spoken": "の"
},
{
"written": "訪日",
"confidence": 1.00,
"starttime": 94532,
"endtime": 95444,
"spoken": "ほうにち"
},
{
"written": "、",
"confidence": 0.68,
"starttime": 95444,
"endtime": 95492,
"spoken": "_"
}
],
"confidence": 0.970,
"starttime": 87700,
"endtime": 95492,
"tags": [],
"rulename": "",
"text": "お答えをいたしました現下の国際情勢のもとで先般のトランプ大統領委員の訪日、"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183859",
"text": "お答えをいたしました現下の国際情勢のもとで先般のトランプ大統領委員の訪日、",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "に",
"confidence": 0.95,
"starttime": 95788,
"endtime": 95980,
"spoken": "に"
},
{
"written": "続き",
"confidence": 1.00,
"starttime": 95980,
"endtime": 96364,
"spoken": "つづき"
},
{
"written": "まし",
"confidence": 1.00,
"starttime": 96364,
"endtime": 96636,
"spoken": "まし"
},
{
"written": "て",
"confidence": 1.00,
"starttime": 96636,
"endtime": 97132,
"spoken": "て"
},
{
"written": "日米",
"confidence": 0.70,
"starttime": 97996,
"endtime": 98460,
"spoken": "にちべい"
},
{
"written": "間",
"confidence": 0.70,
"starttime": 98460,
"endtime": 98684,
"spoken": "かん"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 98684,
"endtime": 98876,
"spoken": "の"
},
{
"written": "緊密",
"confidence": 1.00,
"starttime": 98940,
"endtime": 99644,
"spoken": "きんみつ"
},
{
"written": "な",
"confidence": 1.00,
"starttime": 99644,
"endtime": 100092,
"spoken": "な"
},
{
"written": "連携",
"confidence": 1.00,
"starttime": 100124,
"endtime": 100668,
"spoken": "れんけい"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 100668,
"endtime": 101020,
"spoken": "を"
},
{
"written": "確認",
"confidence": 1.00,
"starttime": 101052,
"endtime": 101500,
"spoken": "かくにん"
},
{
"written": "でき",
"confidence": 0.97,
"starttime": 101500,
"endtime": 101740,
"spoken": "でき"
},
{
"written": "た",
"confidence": 0.99,
"starttime": 101740,
"endtime": 101852,
"spoken": "た"
},
{
"written": "と",
"confidence": 1.00,
"starttime": 101852,
"endtime": 102140,
"spoken": "と"
}
],
"confidence": 0.999,
"starttime": 95500,
"endtime": 102284,
"tags": [],
"rulename": "",
"text": "に続きまして日米間の緊密な連携を確認できたと"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183900",
"text": "に続きまして日米間の緊密な連携を確認できたと",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "思い",
"confidence": 1.00,
"starttime": 102652,
"endtime": 102924,
"spoken": "おもい"
},
{
"written": "ます",
"confidence": 1.00,
"starttime": 102924,
"endtime": 103260,
"spoken": "ます"
},
{
"written": "。",
"confidence": 0.89,
"starttime": 103260,
"endtime": 103324,
"spoken": "_"
},
{
"written": "トランプ大統領",
"confidence": 0.99,
"starttime": 103692,
"endtime": 104636,
"spoken": "とらんぷだいとうりょう"
},
{
"written": "から",
"confidence": 1.00,
"starttime": 104636,
"endtime": 104924,
"spoken": "から"
},
{
"written": "は",
"confidence": 1.00,
"starttime": 104924,
"endtime": 105180,
"spoken": "は"
},
{
"written": "私",
"confidence": 1.00,
"starttime": 105484,
"endtime": 105916,
"spoken": "わたくし"
},
{
"written": "ども",
"confidence": 0.80,
"starttime": 105916,
"endtime": 106060,
"spoken": "ども"
},
{
"written": "は",
"confidence": 0.99,
"starttime": 106060,
"endtime": 106284,
"spoken": "は"
},
{
"written": "極めて",
"confidence": 1.00,
"starttime": 106284,
"endtime": 107132,
"spoken": "きわめて"
},
{
"written": "親しい",
"confidence": 1.00,
"starttime": 107324,
"endtime": 107852,
"spoken": "したしい"
},
{
"written": "友人",
"confidence": 1.00,
"starttime": 107852,
"endtime": 108268,
"spoken": "ゆうじん"
},
{
"written": "で",
"confidence": 1.00,
"starttime": 108268,
"endtime": 108380,
"spoken": "で"
},
{
"written": "あり",
"confidence": 1.00,
"starttime": 108380,
"endtime": 108780,
"spoken": "あり"
},
{
"written": "、",
"confidence": 0.87,
"starttime": 108780,
"endtime": 108844,
"spoken": "_"
}
],
"confidence": 0.979,
"starttime": 102300,
"endtime": 108844,
"tags": [],
"rulename": "",
"text": "思います。トランプ大統領からは私どもは極めて親しい友人であり、"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183901",
"text": "思います。トランプ大統領からは私どもは極めて親しい友人であり、",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "いつ",
"confidence": 0.92,
"starttime": 109058,
"endtime": 109298,
"spoken": "いつ"
},
{
"written": "でも",
"confidence": 0.91,
"starttime": 109298,
"endtime": 109810,
"spoken": "でも"
},
{
"written": "電話",
"confidence": 1.00,
"starttime": 110274,
"endtime": 110658,
"spoken": "でんわ"
},
{
"written": "を",
"confidence": 1.00,
"starttime": 110658,
"endtime": 110754,
"spoken": "を"
},
{
"written": "して",
"confidence": 1.00,
"starttime": 110754,
"endtime": 110946,
"spoken": "して"
},
{
"written": "きて",
"confidence": 1.00,
"starttime": 110946,
"endtime": 111186,
"spoken": "きて"
},
{
"written": "ほしい",
"confidence": 0.82,
"starttime": 111186,
"endtime": 111602,
"spoken": "ほしい"
},
{
"written": "と",
"confidence": 1.00,
"starttime": 111602,
"endtime": 112018,
"spoken": "と"
},
{
"written": "いう",
"confidence": 1.00,
"starttime": 112018,
"endtime": 112242,
"spoken": "いう"
},
{
"written": "お話",
"confidence": 1.00,
"starttime": 112274,
"endtime": 112850,
"spoken": "おはなし"
},
{
"written": "が",
"confidence": 1.00,
"starttime": 112850,
"endtime": 113106,
"spoken": "が"
},
{
"written": "ございました",
"confidence": 1.00,
"starttime": 113138,
"endtime": 113970,
"spoken": "ございました"
},
{
"written": "。",
"confidence": 0.89,
"starttime": 113970,
"endtime": 114194,
"spoken": "_"
}
],
"confidence": 0.999,
"starttime": 108850,
"endtime": 114194,
"tags": [],
"rulename": "",
"text": "いつでも電話をしてきてほしいというお話がございました。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183902",
"text": "いつでも電話をしてきてほしいというお話がございました。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "会談",
"confidence": 1.00,
"starttime": 114584,
"endtime": 115128,
"spoken": "かいだん"
},
{
"written": "内容",
"confidence": 1.00,
"starttime": 115128,
"endtime": 115464,
"spoken": "ないよう"
},
{
"written": "で",
"confidence": 1.00,
"starttime": 115464,
"endtime": 115576,
"spoken": "で"
},
{
"written": "ございます",
"confidence": 1.00,
"starttime": 115576,
"endtime": 116184,
"spoken": "ございます"
},
{
"written": "けれども",
"confidence": 1.00,
"starttime": 116184,
"endtime": 116872,
"spoken": "けれども"
},
{
"written": "これ",
"confidence": 1.00,
"starttime": 117176,
"endtime": 117848,
"spoken": "これ"
},
{
"written": "外交",
"confidence": 1.00,
"starttime": 118232,
"endtime": 118600,
"spoken": "がいこう"
},
{
"written": "上",
"confidence": 1.00,
"starttime": 118600,
"endtime": 118808,
"spoken": "じょう"
},
{
"written": "の",
"confidence": 1.00,
"starttime": 118808,
"endtime": 119096,
"spoken": "の"
},
{
"written": "やり取り",
"confidence": 1.00,
"starttime": 119128,
"endtime": 119640,
"spoken": "やりとり"
},
{
"written": "で",
"confidence": 1.00,
"starttime": 119640,
"endtime": 119736,
"spoken": "で"
},
{
"written": "ございます",
"confidence": 1.00,
"starttime": 119736,
"endtime": 120312,
"spoken": "ございます"
},
{
"written": "ので",
"confidence": 1.00,
"starttime": 120312,
"endtime": 120600,
"spoken": "ので"
},
{
"written": "、",
"confidence": 0.81,
"starttime": 120600,
"endtime": 120632,
"spoken": "_"
},
{
"written": "詳細",
"confidence": 1.00,
"starttime": 120632,
"endtime": 121096,
"spoken": "しょうさい"
},
{
"written": "について",
"confidence": 1.00,
"starttime": 121096,
"endtime": 121592,
"spoken": "について"
},
{
"written": "は",
"confidence": 1.00,
"starttime": 121592,
"endtime": 121816,
"spoken": "は"
}
],
"confidence": 0.999,
"starttime": 114200,
"endtime": 121944,
"tags": [],
"rulename": "",
"text": "会談内容でございますけれどもこれ外交上のやり取りでございますので、詳細については"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183903",
"text": "会談内容でございますけれどもこれ外交上のやり取りでございますので、詳細については",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "差し",
"confidence": 1.00,
"starttime": 121982,
"endtime": 122382,
"spoken": "さし"
},
{
"written": "控え",
"confidence": 1.00,
"starttime": 122382,
"endtime": 122670,
"spoken": "ひかえ"
},
{
"written": "させ",
"confidence": 1.00,
"starttime": 122670,
"endtime": 122958,
"spoken": "させ"
},
{
"written": "て",
"confidence": 1.00,
"starttime": 122958,
"endtime": 123166,
"spoken": "て"
},
{
"written": "ください",
"confidence": 1.00,
"starttime": 123166,
"endtime": 123790,
"spoken": "ください"
},
{
"written": "。",
"confidence": 0.89,
"starttime": 123790,
"endtime": 124046,
"spoken": "_"
}
],
"confidence": 1.000,
"starttime": 121950,
"endtime": 124046,
"tags": [],
"rulename": "",
"text": "差し控えさせてください。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183904",
"text": "差し控えさせてください。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "以上",
"confidence": 1.00,
"starttime": 125142,
"endtime": 125318,
"spoken": "いじょう"
},
{
"written": "で",
"confidence": 0.99,
"starttime": 125318,
"endtime": 125382,
"spoken": "で"
},
{
"written": "終わり",
"confidence": 0.98,
"starttime": 125382,
"endtime": 125558,
"spoken": "おわり"
},
{
"written": "ます",
"confidence": 0.99,
"starttime": 125558,
"endtime": 126374,
"spoken": "ます"
},
{
"written": "。",
"confidence": 0.81,
"starttime": 126374,
"endtime": 126438,
"spoken": "_"
}
],
"confidence": 0.991,
"starttime": 124550,
"endtime": 126438,
"tags": [],
"rulename": "",
"text": "以上で終わります。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183904_",
"text": "以上で終わります。",
"code": "",
"message": ""
}
{
"results": [
{
"tokens": [
{
"written": "お疲れ様です",
"confidence": 1.00,
"starttime": 127212,
"endtime": 128316,
"spoken": "おつかれさまです"
},
{
"written": "。",
"confidence": 0.90,
"starttime": 128316,
"endtime": 128428,
"spoken": "_"
}
],
"confidence": 1.000,
"starttime": 127100,
"endtime": 128588,
"tags": [],
"rulename": "",
"text": "お疲れ様です。"
}
],
"utteranceid": "20260116/18/019bc62be4af0a3035e58536_20260116_183905",
"text": "お疲れ様です。",
"code": "",
"message": ""
}
STEP 3 Subtitling and video composition: Embed subtitle text into the video track
Now we come to the climax. We will create an srt file containing timestamps for the subtitles. This batch file will run json2srt.py. The final subtitle file, "takaichi_voice.srt", will be output. This will be executed as a batch file.
run_jimaku.bat takaichi_voice.json
run_jimaku.bat
@echo off
python json2srt2.py %1
@pauseSubtitle file output
json2srt.py
import argparse
import json
parser = argparse.ArgumentParser()
parser.add_argument("file", help="Designate JSON file name to read")
parser.add_argument("-d", "--delimiters", help="Designate delimiters to separate subtitles. Default value is ['。','、']", default="。,、")
parser.add_argument("-s", "--skip", help="Designate skip words which do not inculud in subtitles. Default value is ['。','、']", default="。,、")
parser.add_argument("-t", "--time", help="Designate allowed time for single subtile by millisecongds. Default value is 5000", default=5000, type=int)
parser.add_argument("-c", "--charas", help="Designate allowed charas for single subtile. Default value is 25", default=25, type=int)
class SRTFomat():
def __init__(self, args):
self.text = ""
self.blocks = []
self.delimiters = args.delimiters.split(",")
self.skipWords = args.skip.split(",")
self.time = args.time
self.charas = args.charas
def readFile(self, file):
f = open(file, "r", encoding="utf-8")
contents = f.read()
f.close()
all_tokens = []
full_text = ""
# Method 1: Try to parse as single JSON object (original format)
try:
data = json.loads(contents)
# Handle original format: {"results": [0]: {"tokens": [], "text": ""}}
if "results" in data:
results = data["results"]
if isinstance(results, list) and len(results) > 0:
# Multiple results in array
for result in results:
if "tokens" in result:
all_tokens.extend(result["tokens"])
if "text" in result:
full_text += result["text"]
elif isinstance(results, dict):
# Single result object
if "tokens" in results:
all_tokens = results["tokens"]
if "text" in results:
full_text = results["text"]
# Direct format without results wrapper
elif "tokens" in data:
all_tokens = data["tokens"]
if "text" in data:
full_text = data["text"]
if all_tokens:
self.text = full_text
self.readTokens(all_tokens)
return
except json.JSONDecodeError:
pass # Continue to try other formats
# Method 2: Try JSONL format (multiple JSON objects per line)
lines = contents.strip().split('\n')
for line in lines:
line = line.strip()
if not line:
continue
try:
data = json.loads(line)
if "results" in data and isinstance(data["results"], list):
for result in data["results"]:
if "tokens" in result:
all_tokens.extend(result["tokens"])
if "text" in result:
full_text += result["text"]
except json.JSONDecodeError:
continue # Skip invalid lines
# Method 3: Try parsing as multi-line JSON objects
if not all_tokens:
current_json = ""
brace_count = 0
for line in lines:
line = line.strip()
if not line:
continue
current_json += line
# Count braces to determine if JSON object is complete
for char in line:
if char == '{':
brace_count += 1
elif char == '}':
brace_count -= 1
# If braces are balanced, we have a complete JSON object
if brace_count == 0 and current_json:
try:
data = json.loads(current_json)
if "results" in data and isinstance(data["results"], list):
for result in data["results"]:
if "tokens" in result:
all_tokens.extend(result["tokens"])
if "text" in result:
full_text += result["text"]
current_json = ""
except json.JSONDecodeError:
current_json = ""
brace_count = 0
if not all_tokens:
print(f"No valid JSON data found in file: {file}")
return
self.text = full_text
self.readTokens(all_tokens)
def readTokens(self, tokens):
sub = ""
startTime = 0
index = 1
# subTitles = []
for token in tokens:
written = token["written"]
# 字幕が空の場合 startTime を設定する
if sub == "":
# 字幕が空でもTokenの中身が句読点などであった場合はスキップ
if written in self.delimiters or written in self.skipWords:
continue
else:
startTime = token["starttime"]
# 字幕区切りをつくっていく
# 各条件で字幕を blocks に格納して一度リセットする
# 句読点にあたったら
if written in self.delimiters or len(sub) > self.charas or token["endtime"] - startTime > self.time:
self.blocks.append(self.createSRTBlock(index, startTime, token["endtime"], sub))
sub = ""
startTime = 0
index += 1
# 条件以外では字幕をつなげていく
else:
if written not in self.skipWords:
sub += token["written"]
# Forループここまで
# 最後のブロックを格納する
self.blocks.append(self.createSRTBlock(index, startTime, tokens[-1]["endtime"], sub))
def createSRTBlock(self, index, startTime, endTime, sub):
stime = self.timeFormat(startTime)
etime = self.timeFormat(endTime)
return f"{index}\n{stime} --> {etime}\n{sub}\n"
def timeFormat(self, time):
time_ = time
ms_ = int(time_ % 1000)
time_ = int((time_ - ms_) / 1000)
sec_ = int(time_ % 60)
time_ = int((time_ - sec_) / 60)
mn_ = int(time_ % 60)
time_ = int((time_ - mn_) /60)
hr_ = int(time_ % 60)
if ms_ < 10:
ms = f"00{ms_}"
elif ms_ < 100:
ms = f"0{ms_}"
else:
ms = str(ms_)
if sec_ < 10:
sec = f"0{sec_}"
else:
sec = str(sec_)
if mn_ < 10:
mn = f"0{mn_}"
else:
mn = str(mn_)
if hr_ < 10:
hr = f"0{hr_}"
else:
hr = str(hr_)
return f"{hr}:{mn}:{sec},{ms}"
def exportSRTText(self):
return "\n".join(self.blocks)
if __name__ == "__main__":
args = parser.parse_args()
if not args.file.endswith(".json"):
print("Please set json file")
else:
srt = SRTFomat(args)
srt.readFile(args.file)
text = srt.exportSRTText()
srtName = args.file.replace(".json", ".srt")
f = open(srtName, "w", encoding="utf-8")
f.write(text)
f.close()
print("done")
The completed subtitle file.
takaichi_voice.srt
1
00:00:00,422 --> 00:00:03,410
皆様どうぞ朝日新聞陣野と申します
2
00:00:03,410 --> 00:00:04,418
よろしくお願いします
3
00:00:04,418 --> 00:00:05,634
よろしくお願いします
4
00:00:05,634 --> 00:00:09,346
総理はアメリカのトランプ大統領と電話会談に臨まれると
5
00:00:09,346 --> 00:00:10,290
が出ています
6
00:00:10,332 --> 00:00:14,092
こちらの実施状況とどちらから持ちかけられたのかについて
7
00:00:14,092 --> 00:00:14,908
します
8
00:00:14,908 --> 00:00:15,372
また
9
00:00:15,372 --> 00:00:16,844
既に実施済みの場合は
10
00:00:16,844 --> 00:00:20,830
会談でどのような議論をされたのでしょうか？台湾問題や
11
00:00:20,830 --> 00:00:23,390
の存立危機事態を巡る答弁について
12
00:00:23,390 --> 00:00:26,766
どのようなやり取りをされたのかについてもあわせて伺い
13
00:00:27,338 --> 00:00:27,994
はい
14
00:00:28,112 --> 00:00:29,376
ありがとうございます
15
00:00:29,664 --> 00:00:34,608
先ほどトランプ大統領からのお申し出によりまして電話会談
16
00:00:34,608 --> 00:00:35,584
行いました
17
00:00:36,144 --> 00:00:41,344
私からはトランプ大統領が訪日されたときのことを
18
00:00:41,344 --> 00:00:41,936
した
19
00:00:42,526 --> 00:00:47,758
アメリカの新聞に採用したのを送ってきてくださったこと
20
00:00:47,758 --> 00:00:53,078
お礼とあとウクライナ和平に関する米国の
21
00:00:53,078 --> 00:00:58,196
への評価についてお伝えをしましたその上で
22
00:00:58,196 --> 00:00:59,492
同盟の強化や
23
00:00:59,612 --> 00:01:04,732
インド太平洋地域が直面する情勢や諸課題について幅広く
24
00:01:04,732 --> 00:01:06,444
交換を行いました
25
00:01:06,738 --> 00:01:07,474
その中で
26
00:01:07,474 --> 00:01:13,106
トランプ大統領からは昨晩行われた米中首脳会談を
27
00:01:13,550 --> 00:01:19,326
最近の米中関係の状況について説明が
28
00:01:19,578 --> 00:01:24,842
あわせて私が出席したGTPの
29
00:01:24,842 --> 00:01:27,690
についても尋ねられましたので
30
00:01:27,764 --> 00:01:33,092
お答えをいたしました現下の国際情勢のもとで
31
00:01:33,092 --> 00:01:35,492
のトランプ大統領委員の訪日
32
00:01:35,788 --> 00:01:41,020
に続きまして日米間の緊密な連携
33
00:01:41,052 --> 00:01:43,324
確認できたと思います
34
00:01:43,692 --> 00:01:48,780
トランプ大統領からは私どもは極めて親しい友人で
35
00:01:49,058 --> 00:01:54,194
いつでも電話をしてきてほしいというお話がございました
36
00:01:54,584 --> 00:01:59,640
会談内容でございますけれどもこれ外交上の
37
00:01:59,640 --> 00:02:00,632
でございますので
38
00:02:00,632 --> 00:02:04,046
詳細については差し控えさせてください
39
00:02:05,142 --> 00:02:06,438
以上で終わります
40
00:02:07,212 --> 00:02:08,428
お疲れ様です
41
00:00:00,000 --> 00:02:08,428
This is the final step. Place takaichi_voice.srt in the same folder as the video. The key here is to give the video mp4 file and the srt file the same file name. This will ensure that the srt file is loaded correctly.
FINAL Watch the video and read the subtitles
When you open the video "takaichi_voice.mp4" in VLC media player, you should be able to select track 1 from the subtitle tracks.
And then, the subtitles appeared!
The timing of the subtitles was perfect.
To the end
In the past, programs labeled as subtitled, data-based, or bilingual in the TV guide in newspapers were considered special and valuable. They were thought to have required large production costs.
However, thanks to advances in voice recognition technology and image software, we are now in an age where subtitles can be easily created. By utilizing generative AI, it is possible to add new value to textual information by translating, paraphrasing, summarizing, and so on.
I hope that the number of Vloggers will continue to grow and that many fun videos will be created. The possibilities brought about by new technology are endless. I hope you will all join in and enjoy this wave.
Until the end Thank you for reading.
*This article is intended to explain the technical workflow. For videos you actually use, please follow the terms of use and copyright of each distribution source or platform. We recommend viewing or embedding the Prime Minister's Office videos on their official page or official YouTube channel.
Person who wrote this article
-
Spice Dog
A brown-colored engineer with a deep love for Shiba Inu dogs and Indian curry, he decided to take a fresh start and dive into the world of voice recognition, driven by a desire to add some spice to his daily life.
Recently, I noticed that my dog can distinguish the engine sounds of Subaru cars, and I've been secretly daydreaming about whether a dog's hearing ability could be used in voice recognition technology.
Most viewed articles
-
A quick explanation of how speech recognition works!
-
Comparing the speech recognition rates of OpenAI's Whisper and AmiVoice for "conference" audio
-
How to use the AmiVoice API free coupon
New columns
-
AmiVoice API Update Explanation: New Parameters for Voicebots Reduce Response Wait Times
-
AmiVoice API Update: End-to-End ASR–Ready ”Keyword Biasing”
-
Easily synthesize subtitles into videos! Subtitle workflow created with speech recognition API
Category list
- Introduction to Speech Recognition (15)
- How to improve voice recognition accuracy (12)
- I tried developing it (27)
- How to use AmiVoiceAPI(26)
- Comparison and Verification (5)
- Others(9)
Archive
