update:音色克隆

This commit is contained in:
rainv123
2025-10-13 14:17:53 +08:00
parent 99f178274d
commit 087fd5d547
3 changed files with 118 additions and 1 deletions
@@ -175,7 +175,9 @@ public class VoiceCloneController {
public Result<String> cloneAudio(@RequestBody Map<String, String> params) {
String cloneId = params.get("cloneId");
checkPermission(cloneId);
return new Result<String>();
// 调用服务层进行语音克隆训练
String result = voiceCloneService.cloneAudio(cloneId);
return new Result<String>().ok(result);
}
private void checkPermission(String id) {
@@ -65,4 +65,11 @@ public interface VoiceCloneService extends BaseService<VoiceCloneEntity> {
* 获取音频数据
*/
byte[] getVoiceData(String id);
/**
* 克隆音频,调用火山引擎进行语音复刻训练
* @param cloneId 语音克隆记录ID
* @return 训练结果消息
*/
String cloneAudio(String cloneId);
}
@@ -1,7 +1,14 @@
package xiaozhi.modules.voiceclone.service.impl;
import java.io.IOException;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;
import java.net.URI;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Base64;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
@@ -12,14 +19,17 @@ import org.springframework.web.multipart.MultipartFile;
import com.baomidou.mybatisplus.core.conditions.query.QueryWrapper;
import com.baomidou.mybatisplus.core.metadata.IPage;
import com.fasterxml.jackson.databind.ObjectMapper;
import lombok.RequiredArgsConstructor;
import lombok.extern.slf4j.Slf4j;
import xiaozhi.common.page.PageData;
import xiaozhi.common.service.impl.BaseServiceImpl;
import xiaozhi.common.utils.ConvertUtils;
import xiaozhi.common.utils.DateUtils;
import xiaozhi.common.exception.ErrorCode;
import xiaozhi.common.exception.RenException;
import xiaozhi.modules.model.entity.ModelConfigEntity;
import xiaozhi.modules.model.service.ModelConfigService;
import xiaozhi.modules.sys.service.SysUserService;
import xiaozhi.modules.voiceclone.dao.VoiceCloneDao;
@@ -28,6 +38,7 @@ import xiaozhi.modules.voiceclone.dto.VoiceCloneResponseDTO;
import xiaozhi.modules.voiceclone.entity.VoiceCloneEntity;
import xiaozhi.modules.voiceclone.service.VoiceCloneService;
@Slf4j
@Service
@RequiredArgsConstructor
public class VoiceCloneServiceImpl extends BaseServiceImpl<VoiceCloneDao, VoiceCloneEntity>
@@ -35,6 +46,7 @@ public class VoiceCloneServiceImpl extends BaseServiceImpl<VoiceCloneDao, VoiceC
private final ModelConfigService modelConfigService;
private final SysUserService sysUserService;
private final ObjectMapper objectMapper;
@Override
public PageData<VoiceCloneEntity> page(Map<String, Object> params) {
@@ -207,4 +219,100 @@ public class VoiceCloneServiceImpl extends BaseServiceImpl<VoiceCloneDao, VoiceC
}
return entity.getVoice();
}
@Override
@Transactional(rollbackFor = Exception.class)
public String cloneAudio(String cloneId) {
VoiceCloneEntity entity = baseDao.selectById(cloneId);
if (entity == null) {
throw new RenException(ErrorCode.VOICE_CLONE_RECORD_NOT_EXIST);
}
if (entity.getVoice() == null || entity.getVoice().length == 0) {
throw new RenException("请先上传音频文件");
}
try {
ModelConfigEntity modelConfig =
modelConfigService.getModelByIdFromCache("TTS_HuoshanDoubleStreamTTS");
if (modelConfig == null || modelConfig.getConfigJson() == null) {
throw new RenException("火山引擎配置未找到");
}
Map<String, Object> config = modelConfig.getConfigJson();
String appid = (String) config.get("appid");
String accessToken = (String) config.get("access_token");
if (StringUtils.isAnyBlank(appid, accessToken)) {
throw new RenException("火山引擎配置不完整");
}
String audioBase64 = Base64.getEncoder().encodeToString(entity.getVoice());
Map<String, Object> reqBody = new HashMap<>();
reqBody.put("appid", appid);
reqBody.put("name", entity.getName());
List<Map<String, String>> audios = new ArrayList<>();
Map<String, String> audioMap = new HashMap<>();
audioMap.put("audio_bytes", audioBase64);
audioMap.put("audio_format", "wav");
audios.add(audioMap);
reqBody.put("audios", audios);
reqBody.put("source", 2);
reqBody.put("language", 0);
reqBody.put("model_type", 1);
reqBody.put("speaker_id", entity.getName());
String apiUrl = "https://openspeech.bytedance.com/api/v1/mega_tts/audio/upload";
HttpClient client = HttpClient.newHttpClient();
HttpRequest request = HttpRequest.newBuilder()
.uri(URI.create(apiUrl))
.header("Content-Type", "application/json")
.header("Authorization", "Bearer;" + accessToken)
.header("Resource-Id", "volc.megatts.voiceclone")
.POST(HttpRequest.BodyPublishers.ofString(objectMapper.writeValueAsString(reqBody)))
.build();
// >>> 打印调试信息
System.out.println(">>> 请求地址 = " + apiUrl);
System.out.println(">>> AppID = " + appid);
System.out.println(">>> Token = " + accessToken);
/* ===== 真正发请求 ===== */
HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
System.out.println(">>> HTTP status = " + response.statusCode());
System.out.println(">>> response body = " + response.body());
if (response.statusCode() == 200) {
Map<String, Object> rsp = objectMapper.readValue(response.body(), Map.class);
String status = (String) rsp.get("status");
String spkId = (String) rsp.get("spk_id");
if ("success".equals(status) && StringUtils.isNotBlank(spkId)) {
entity.setTrainStatus(2);
entity.setVoiceId(spkId);
entity.setTrainError(null);
baseDao.updateById(entity);
return "训练成功,声音ID: " + spkId;
} else {
String msg = (String) rsp.getOrDefault("message", "训练失败");
throw new RenException(msg);
}
} else {
System.out.println(">>> 请求失败 HTTP = " + response.statusCode());
System.out.println(">>> 失败 body = " + response.body());
throw new RenException("请求失败,状态码: " + response.statusCode());
}
} catch (RenException re) {
// 自己抛的业务异常继续往外抛
throw re;
} catch (Exception e) {
/* ===== 任何其他异常(空指针、IO、JSON、404...)全部打印 ===== */
System.out.println(">>> 语音克隆异常: " + e.getMessage());
e.printStackTrace(); // 详细堆栈
entity.setTrainStatus(3);
entity.setTrainError(e.getMessage());
baseDao.updateById(entity);
throw new RenException("语音克隆失败: " + e.getMessage());
}
}
}