[logger/test]添加debug级日志记录接收数据长度，测试说话人识别效果。

src/functor/asr_functor.py

@@ -112,6 +112,9 @@ class ASRFunctor(BaseFunctor):
         while self._is_running:
             try:
                 data = self._input_queue.get(True, timeout=1)
+                if data is None:
+                    break
+                logger.debug("[ASRFunctor]获取到的数据length: %s", len(data))
                 self._process(data)
                 self._input_queue.task_done()
             # 当队列为空时, 间隔1s检测是否进入停止事件。

src/functor/spk_functor.py

@@ -98,6 +98,9 @@ class SPKFunctor(BaseFunctor):
         while self._is_running:
             try:
                 data = self._input_queue.get(True, timeout=1)
+                if data is None:
+                    break
+                logger.debug("[SPKFunctor]获取到的数据length: %s", len(data))
                 self._process(data)
                 self._input_queue.task_done()
             # 当队列为空时, 间隔1s检测是否进入停止事件。

src/functor/vad_functor.py

@@ -202,6 +202,9 @@ class VADFunctor(BaseFunctor):
         while self._is_running:
             try:
                 data = self._input_queue.get(True, timeout=1)
+                if data is None:
+                    break
+                logger.debug("[VADFunctor]获取到的数据length: %s", len(data))
                 self._process(data)
                 self._input_queue.task_done()
             # 当队列为空时, 间隔1s检测是否进入停止事件。

src/pipeline/ASRpipeline.py

@@ -224,6 +224,7 @@ class ASRPipeline(PipelineBase):
         while self._is_running and not self._stop_event:
             try:
                 data = self._input_queue.get(timeout=self._queue_timeout)
+                logger.debug("[ASRpipeline]获取到的数据length: %s", len(data))
                 # 检查是否是结束信号
                 if data is None:
                     logger.info("收到结束信号，管道准备停止")

tests/spkverify_use.py

@@ -0,0 +1,23 @@
+from modelscope.pipelines import pipeline
+sv_pipeline = pipeline(
+    task='speaker-verification',
+    model='iic/speech_campplus_sv_zh-cn_16k-common',
+    model_revision='v1.0.0'
+)
+speaker1_a_wav = 'https://modelscope.cn/api/v1/models/damo/speech_campplus_sv_zh-cn_16k-common/repo?Revision=master&FilePath=examples/speaker1_a_cn_16k.wav'
+speaker1_b_wav = 'https://modelscope.cn/api/v1/models/damo/speech_campplus_sv_zh-cn_16k-common/repo?Revision=master&FilePath=examples/speaker1_b_cn_16k.wav'
+speaker2_a_wav = 'https://modelscope.cn/api/v1/models/damo/speech_campplus_sv_zh-cn_16k-common/repo?Revision=master&FilePath=examples/speaker2_a_cn_16k.wav'
+# 相同说话人语音
+result = sv_pipeline([speaker1_a_wav, speaker1_b_wav])
+print(result)
+# 不同说话人语音
+result = sv_pipeline([speaker1_a_wav, speaker2_a_wav])
+print(result)
+# 可以自定义得分阈值来进行识别，阈值越高，判定为同一人的条件越严格
+result = sv_pipeline([speaker1_a_wav, speaker1_a_wav], thr=0.6)
+print(result)
+# 可以传入output_emb参数，输出结果中就会包含提取到的说话人embedding
+result = sv_pipeline([speaker1_a_wav, speaker2_a_wav], output_emb=True)
+print(result['embs'], result['outputs'])
+# 可以传入save_dir参数，提取到的说话人embedding会存储在save_dir目录中
+result = sv_pipeline([speaker1_a_wav, speaker2_a_wav], save_dir='savePath/')