一、评测模式描述
·评测要求:支持多组评估文本,每组不超过30个字。音频数据最长60秒。
·评测维度:支持返回单词精准度,流利度,完整度;支持返回音素精准度。
·评测功能:多组文本,音素到字母映射,音素到国际音标。
二、请求参数
主要请求参数说明:
| 参数名称 | 类型 | 描述 |
| ref_text | String | 被评估文本。可以使用 | 划分成多个分支 |
| eval_mode | Integer | 评估模式。6:句子多分支评测模式 |
请求示例
三、返回结果
主要返回结果说明:
| 参数名称 | 类型 | 描述 |
| SuggestedScore | Float | 建议评分 |
| PronAccuracy | Float | 整体精准度 |
| PronFluency | Float | 整体流利度 |
| PronCompletion | Float | 整体完整度 |
| Words.PronAccuracy | Float | 单词精准度 |
| Words.PronFluency | Float | 单词流利度 |
| Words.MatchTag | Integer | 当前词的音频与文本的匹配情况 |
| Words.PhoneInfos.PronAccuracy | Float | 音素精准度 |
| Words.PhoneInfos.MatchTag | Integer | 当前音素的音频与文本的匹配情况 |
| RefTextId | Integer | 匹配候选文本的序号 |
返回示例
{
"code": 0,
"message": "7ed2d079-1e81-4bcc-be4b-45078d10c77c_11",
"voice_id": "7ed2d079-1e81-4bcc-be4b-45078d10c77c",
"result": {
"SuggestedScore": 98.80253601074219,
"PronAccuracy": 98.80253601074219,
"PronFluency": 0.9656875133514404,
"PronCompletion": 1,
"Words": [
{
"MemBeginTime": 210,
"MemEndTime": 350,
"PronAccuracy": 96.01470947265625,
"PronFluency": 0.9704024791717529,
"ReferenceWord": "i_0",
"Word": "i",
"MatchTag": 0,
"KeywordTag": 0,
"PhoneInfos": [
{
"MemBeginTime": 210,
"MemEndTime": 350,
"PronAccuracy": 96.01470947265625,
"DetectedStress": false,
"Phone": "ay",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
}
],
"Tone": {
"Valid": false,
"RefTone": -1,
"HypothesisTone": -1
}
},
{
"MemBeginTime": 350,
"MemEndTime": 530,
"PronAccuracy": 99.027587890625,
"PronFluency": 0.9832581281661987,
"ReferenceWord": "go_1",
"Word": "go",
"MatchTag": 0,
"KeywordTag": 0,
"PhoneInfos": [
{
"MemBeginTime": 350,
"MemEndTime": 430,
"PronAccuracy": 99.00126647949219,
"DetectedStress": false,
"Phone": "g",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 430,
"MemEndTime": 530,
"PronAccuracy": 99.05391693115234,
"DetectedStress": false,
"Phone": "ow",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
}
],
"Tone": {
"Valid": false,
"RefTone": -1,
"HypothesisTone": -1
}
},
{
"MemBeginTime": 530,
"MemEndTime": 630,
"PronAccuracy": 98.93866729736328,
"PronFluency": 1,
"ReferenceWord": "to_2",
"Word": "to",
"MatchTag": 0,
"KeywordTag": 0,
"PhoneInfos": [
{
"MemBeginTime": 530,
"MemEndTime": 590,
"PronAccuracy": 98.98005676269531,
"DetectedStress": false,
"Phone": "t",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 590,
"MemEndTime": 630,
"PronAccuracy": 98.89727783203125,
"DetectedStress": false,
"Phone": "ah",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
}
],
"Tone": {
"Valid": false,
"RefTone": -1,
"HypothesisTone": -1
}
},
{
"MemBeginTime": 630,
"MemEndTime": 990,
"PronAccuracy": 99.04634094238281,
"PronFluency": 0.9832895994186401,
"ReferenceWord": "school_3",
"Word": "school",
"MatchTag": 0,
"KeywordTag": 0,
"PhoneInfos": [
{
"MemBeginTime": 630,
"MemEndTime": 730,
"PronAccuracy": 99.14324951171875,
"DetectedStress": false,
"Phone": "s",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 730,
"MemEndTime": 810,
"PronAccuracy": 99.23045349121094,
"DetectedStress": false,
"Phone": "k",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 810,
"MemEndTime": 910,
"PronAccuracy": 99.08352661132812,
"DetectedStress": false,
"Phone": "uw",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 910,
"MemEndTime": 990,
"PronAccuracy": 98.7281265258789,
"DetectedStress": false,
"Phone": "l",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
}
],
"Tone": {
"Valid": false,
"RefTone": -1,
"HypothesisTone": -1
}
},
{
"MemBeginTime": 990,
"MemEndTime": 1150,
"PronAccuracy": 99.0534439086914,
"PronFluency": 0.9959675073623657,
"ReferenceWord": "by_4",
"Word": "by",
"MatchTag": 0,
"KeywordTag": 0,
"PhoneInfos": [
{
"MemBeginTime": 990,
"MemEndTime": 1050,
"PronAccuracy": 98.91246795654297,
"DetectedStress": false,
"Phone": "b",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 1050,
"MemEndTime": 1150,
"PronAccuracy": 99.19442749023438,
"DetectedStress": false,
"Phone": "ay",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
}
],
"Tone": {
"Valid": false,
"RefTone": -1,
"HypothesisTone": -1
}
},
{
"MemBeginTime": 1150,
"MemEndTime": 1730,
"PronAccuracy": 98.99871063232422,
"PronFluency": 0.8612076640129089,
"ReferenceWord": "bus_5",
"Word": "bus",
"MatchTag": 0,
"KeywordTag": 0,
"PhoneInfos": [
{
"MemBeginTime": 1150,
"MemEndTime": 1250,
"PronAccuracy": 98.90721893310547,
"DetectedStress": false,
"Phone": "b",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 1250,
"MemEndTime": 1470,
"PronAccuracy": 99.09923553466797,
"DetectedStress": false,
"Phone": "ah",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
},
{
"MemBeginTime": 1470,
"MemEndTime": 1730,
"PronAccuracy": 98.98968505859375,
"DetectedStress": false,
"Phone": "s",
"ReferencePhone": "",
"ReferenceLetter": "",
"Stress": false,
"MatchTag": 0
}
],
"Tone": {
"Valid": false,
"RefTone": -1,
"HypothesisTone": -1
}
}
],
"SentenceId": 0,
"RefTextId": 0,
"KeyWordHits": null,
"UnKeyWordHits": null
},
"final": 1
}
四、音素到字母映射
通过该功能,可以对音素映射字母进行标记。音素到字母映射结构:{::cmd{F_P2L=true}} + 评估文本。
五、请求参数
主要请求参数说明:
| 参数名称 | 类型 | 描述 |
| ref_text | String | 被评估文本 |
| eval_mode | Integer | 评估模式。1:句子评测模式 |
请求示例
# 参数示例为websocket连接URL展开, 如:soe.cloud.tencent.com/soe/api/1306***?eval_mode=0&voice_format=1&…server_engine_type=16k_eneval_mode=6# 长音频推荐流式rec_mode=0# 流式展示中间结果sentence_info_enabled=1ref_text="{::cmd{F_P2L=true}}I would like to eat an orange"score_coeff=1.000000voice_format=1
六、返回结果
主要返回结果说明:
| 参数名称 | 类型 | 描述 |
| Words.Word | String | 当前单词 |
| Words.PhoneInfos.Phone | String | 当前单词的音素 |
| Words.PhoneInfos.ReferenceLetter | String | 当前单词的音素映射的字母 |
返回示例
可以参照句子模式相应示例:句子评测模式
七、音素到国际音标转换
返回音素默认为智聆音素,使用 {::cmd{F_IPA=true}} + 单词 开启音素到国际音标转换功能。
八、请求参数
主要请求参数说明:
| 参数名称 | 类型 | 描述 |
| ref_text | String | 被评估文本 |
| eval_mode | Integer | 评估模式。6:句子多分支评测模式 |
请求示例
# 参数示例为websocket连接URL展开, 如:soe.cloud.tencent.com/soe/api/1306***?eval_mode=0&voice_format=1&…server_engine_type=16k_eneval_mode=6# 长音频推荐流式rec_mode=0# 流式展示中间结果sentence_info_enabled=1ref_text="{::cmd{F_IPA=true}}I would like to eat an orange"score_coeff=1.000000voice_format=1
九、返回结果
返回结果说明:
| 参数名称 | 类型 | 描述 |
| Words.PhoneInfos.Phone | String | 当前音节 |
返回示例
可以参照句子模式相应示例:句子评测模式