From 580b11b57ac4b62f7e2acda73813a4e10e8e4cd3 Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期二, 10 十月 2023 17:17:29 +0800
Subject: [PATCH] v0.8.0
---
funasr/models/e2e_vad.py | 64 +++++++++++++++++++++++++------
1 files changed, 51 insertions(+), 13 deletions(-)
diff --git a/funasr/models/e2e_vad.py b/funasr/models/e2e_vad.py
index 440a049..7c55b2e 100644
--- a/funasr/models/e2e_vad.py
+++ b/funasr/models/e2e_vad.py
@@ -5,6 +5,7 @@
from torch import nn
import math
from funasr.models.encoder.fsmn_encoder import FSMN
+from funasr.models.base_model import FunASRModel
class VadStateMachine(Enum):
@@ -35,6 +36,11 @@
class VADXOptions:
+ """
+ Author: Speech Lab of DAMO Academy, Alibaba Group
+ Deep-FSMN for Large Vocabulary Continuous Speech Recognition
+ https://arxiv.org/abs/1803.05030
+ """
def __init__(
self,
sample_rate: int = 16000,
@@ -99,6 +105,11 @@
class E2EVadSpeechBufWithDoa(object):
+ """
+ Author: Speech Lab of DAMO Academy, Alibaba Group
+ Deep-FSMN for Large Vocabulary Continuous Speech Recognition
+ https://arxiv.org/abs/1803.05030
+ """
def __init__(self):
self.start_ms = 0
self.end_ms = 0
@@ -117,6 +128,11 @@
class E2EVadFrameProb(object):
+ """
+ Author: Speech Lab of DAMO Academy, Alibaba Group
+ Deep-FSMN for Large Vocabulary Continuous Speech Recognition
+ https://arxiv.org/abs/1803.05030
+ """
def __init__(self):
self.noise_prob = 0.0
self.speech_prob = 0.0
@@ -126,6 +142,11 @@
class WindowDetector(object):
+ """
+ Author: Speech Lab of DAMO Academy, Alibaba Group
+ Deep-FSMN for Large Vocabulary Continuous Speech Recognition
+ https://arxiv.org/abs/1803.05030
+ """
def __init__(self, window_size_ms: int, sil_to_speech_time: int,
speech_to_sil_time: int, frame_size_ms: int):
self.window_size_ms = window_size_ms
@@ -191,7 +212,12 @@
return int(self.frame_size_ms)
-class E2EVadModel(nn.Module):
+class E2EVadModel(FunASRModel):
+ """
+ Author: Speech Lab of DAMO Academy, Alibaba Group
+ Deep-FSMN for Large Vocabulary Continuous Speech Recognition
+ https://arxiv.org/abs/1803.05030
+ """
def __init__(self, encoder: FSMN, vad_post_args: Dict[str, Any], frontend=None):
super(E2EVadModel, self).__init__()
self.vad_opts = VADXOptions(**vad_post_args)
@@ -201,7 +227,6 @@
self.vad_opts.frame_in_ms)
self.encoder = encoder
# init variables
- self.is_final = False
self.data_buf_start_frame = 0
self.frm_cnt = 0
self.latest_confirmed_speech_frame = 0
@@ -228,11 +253,10 @@
self.data_buf = None
self.data_buf_all = None
self.waveform = None
- self.ResetDetection()
self.frontend = frontend
+ self.last_drop_frames = 0
def AllResetDetection(self):
- self.is_final = False
self.data_buf_start_frame = 0
self.frm_cnt = 0
self.latest_confirmed_speech_frame = 0
@@ -259,7 +283,8 @@
self.data_buf = None
self.data_buf_all = None
self.waveform = None
- self.ResetDetection()
+ self.last_drop_frames = 0
+ self.windows_detector.Reset()
def ResetDetection(self):
self.continous_silence_frame_count = 0
@@ -271,6 +296,15 @@
self.windows_detector.Reset()
self.sil_frame = 0
self.frame_probs = []
+
+ if self.output_data_buf:
+ assert self.output_data_buf[-1].contain_seg_end_point == True
+ drop_frames = int(self.output_data_buf[-1].end_ms / self.vad_opts.frame_in_ms)
+ real_drop_frames = drop_frames - self.last_drop_frames
+ self.last_drop_frames = drop_frames
+ self.data_buf_all = self.data_buf_all[real_drop_frames * int(self.vad_opts.frame_in_ms * self.vad_opts.sample_rate / 1000):]
+ self.decibel = self.decibel[real_drop_frames:]
+ self.scores = self.scores[:, real_drop_frames:, :]
def ComputeDecibel(self) -> None:
frame_sample_length = int(self.vad_opts.frame_length_ms * self.vad_opts.sample_rate / 1000)
@@ -286,7 +320,7 @@
0.000001))
def ComputeScores(self, feats: torch.Tensor, in_cache: Dict[str, torch.Tensor]) -> None:
- scores = self.encoder(feats, in_cache) # return B * T * D
+ scores = self.encoder(feats, in_cache).to('cpu') # return B * T * D
assert scores.shape[1] == feats.shape[1], "The shape between feats and scores does not match"
self.vad_opts.nn_eval_block_size = scores.shape[1]
self.frm_cnt += scores.shape[1] # count total frames
@@ -299,7 +333,7 @@
while self.data_buf_start_frame < frame_idx:
if len(self.data_buf) >= int(self.vad_opts.frame_in_ms * self.vad_opts.sample_rate / 1000):
self.data_buf_start_frame += 1
- self.data_buf = self.data_buf_all[self.data_buf_start_frame * int(
+ self.data_buf = self.data_buf_all[(self.data_buf_start_frame - self.last_drop_frames) * int(
self.vad_opts.frame_in_ms * self.vad_opts.sample_rate / 1000):]
def PopDataToOutputBuf(self, start_frm: int, frm_cnt: int, first_frm_is_start_point: bool,
@@ -444,10 +478,12 @@
- 1)) / self.vad_opts.noise_frame_num_used_for_snr
return frame_state
-
+
def forward(self, feats: torch.Tensor, waveform: torch.tensor, in_cache: Dict[str, torch.Tensor] = dict(),
is_final: bool = False
) -> Tuple[List[List[List[int]]], Dict[str, torch.Tensor]]:
+ if not in_cache:
+ self.AllResetDetection()
self.waveform = waveform # compute decibel for each frame
self.ComputeDecibel()
self.ComputeScores(feats, in_cache)
@@ -474,11 +510,13 @@
return segments, in_cache
def forward_online(self, feats: torch.Tensor, waveform: torch.tensor, in_cache: Dict[str, torch.Tensor] = dict(),
- is_final: bool = False, max_end_sil: int = 800
- ) -> Tuple[List[List[List[int]]], Dict[str, torch.Tensor]]:
+ is_final: bool = False, max_end_sil: int = 800
+ ) -> Tuple[List[List[List[int]]], Dict[str, torch.Tensor]]:
+ if not in_cache:
+ self.AllResetDetection()
self.max_end_sil_frame_cnt_thresh = max_end_sil - self.vad_opts.speech_to_sil_time_thres
self.waveform = waveform # compute decibel for each frame
-
+
self.ComputeScores(feats, in_cache)
self.ComputeDecibel()
if not is_final:
@@ -516,7 +554,7 @@
return 0
for i in range(self.vad_opts.nn_eval_block_size - 1, -1, -1):
frame_state = FrameState.kFrameStateInvalid
- frame_state = self.GetFrameState(self.frm_cnt - 1 - i)
+ frame_state = self.GetFrameState(self.frm_cnt - 1 - i - self.last_drop_frames)
self.DetectOneFrame(frame_state, self.frm_cnt - 1 - i, False)
return 0
@@ -526,7 +564,7 @@
return 0
for i in range(self.vad_opts.nn_eval_block_size - 1, -1, -1):
frame_state = FrameState.kFrameStateInvalid
- frame_state = self.GetFrameState(self.frm_cnt - 1 - i)
+ frame_state = self.GetFrameState(self.frm_cnt - 1 - i - self.last_drop_frames)
if i != 0:
self.DetectOneFrame(frame_state, self.frm_cnt - 1 - i, False)
else:
--
Gitblit v1.9.1