Ziyi223 commited on
Commit
77be746
·
verified ·
1 Parent(s): d75605b

Update include/ten_vad.h

Browse files
Files changed (1) hide show
  1. include/ten_vad.h +5 -3
include/ten_vad.h CHANGED
@@ -39,7 +39,8 @@ extern "C"
39
  * @param[in] hop_size The number of samples between the start points of
40
  * two consecutive analysis frames. (e.g., 256).
41
  * @param[in] threshold VAD detection threshold ranging from [0.0, 1.0]
42
- * (default: 0.5).
 
43
  * @return 0 on success, or -1 error occurs.
44
  */
45
  TENVAD_API int ten_vad_create(ten_vad_handle_t *handle, size_t hop_size,
@@ -54,9 +55,10 @@ extern "C"
54
  * buffer length must equal the hop size specified at ten_vad_create.
55
  * @param[in] audio_data_length size of audio_data buffer, here should be equal to hop_size.
56
  * @param[out] out_probability Pointer to a float (size 1) that receives the
57
- * voice activity probability in the range [0.0, 1.0].
58
  * @param[out] out_flag Pointer to an int (size 1) that receives the
59
- * detection result: 0 = no voice, 1 = voice detected.
 
60
  * @return 0 on success, or -1 error occurs.
61
  */
62
  TENVAD_API int ten_vad_process(ten_vad_handle_t handle, const int16_t *audio_data, size_t audio_data_length,
 
39
  * @param[in] hop_size The number of samples between the start points of
40
  * two consecutive analysis frames. (e.g., 256).
41
  * @param[in] threshold VAD detection threshold ranging from [0.0, 1.0]
42
+ * This threshold is used to determine voice activity by comparing with the output probability.
43
+ * When probability >= threshold, voice is detected.
44
  * @return 0 on success, or -1 error occurs.
45
  */
46
  TENVAD_API int ten_vad_create(ten_vad_handle_t *handle, size_t hop_size,
 
55
  * buffer length must equal the hop size specified at ten_vad_create.
56
  * @param[in] audio_data_length size of audio_data buffer, here should be equal to hop_size.
57
  * @param[out] out_probability Pointer to a float (size 1) that receives the
58
+ * voice activity probability in the range [0.0, 1.0], where higher values indicate higher confidence in voice presence.
59
  * @param[out] out_flag Pointer to an int (size 1) that receives the
60
+ * binary voice activity decision: 0: no voice, 1: voice detected.
61
+ * This flag is set to 1 when out_probability >= threshold, and 0 otherwise.
62
  * @return 0 on success, or -1 error occurs.
63
  */
64
  TENVAD_API int ten_vad_process(ten_vad_handle_t handle, const int16_t *audio_data, size_t audio_data_length,