vad : add initial Voice Activity Detection (VAD) support #3065
danbev
force pushed
from
0f2fe06f
to
b10e6ddd
1 year ago
danbev
force pushed
from
4aea8b3a
to
3cca1a23
1 year ago
danbev
force pushed
from
5758650f
to
9f0ed3d8
1 year ago
danbev
force pushed
from
63d3fe24
to
ebc79f8a
1 year ago
danbev
force pushed
from
b59768bb
to
798695fc
1 year ago
danbev
marked this pull request as ready for review 1 year ago
danbev
force pushed
from
798695fc
to
6b56b7df
1 year ago
ggerganov
force pushed
from
5a6236e9
to
60d561b8
1 year ago
danbev
commented
on 2025-05-09
vad : add initial Voice Activity Detection (VAD) support
871da0bf
examples : add VAD parameters to CLI [no ci]
24901683
ci : add job to test VAD
eb23253b
vad : map timestamps to original audio
59252c2b
squash! vad : add initial Voice Activity Detection (VAD) support [no ci]
37a36a33
vad : extract VAD processing to a separate function
033c0ce2
vad : add TODOs to optimize segment access [no ci]
028481e9
vad : only use CPU backend for VAD processing [no ci]
fc7ebf20
tests : fix strcmp assert and use beam search
3276232e
vad : dont reshape stft_forward_basis tensor
abc05c5c
vad : use ggml_row_size() and rename hdim_bytes to hdim_size
0e18ceba
vad : remove unnecessary ggml_cont
9bf1b4b3
vad : fix typo in log message
dc529950
vad : don't use left leaning ref for segment
2b057733
vad : use std::vector<float> instead float pointers
44bdef1b
vad : enable GPU support for VAD but default to false
27eb59bd
vad : use kebab-case and not snake_case for VAD options
643a91bf
vad : add h_state and c_state to whisper_vad_state
e4d43072
vad : always initialize filtered_n_samples to 0
94c3aba8
vad : use orig timestamp for first segment
e70e4861
vad : fix buffers and enable GPU support by default
436baeb7
vad : fix use_gpu assert in test-vad.cpp
eb2c83ee
vad : remove unnecessary reserve [no ci]
47c8f02d
vad : add probs to whisper_vad_state
327cdaee
vad : add timing of vad processing [no ci]
bf2b0df9
danbev
force pushed
from
50337e26
to
bf2b0df9
1 year ago
vad : force GPU off for now
243e0dba
vad : minor style and naming changes
65c421de
vad : minor style
cae38fda
vad : remove obsolete whisper_vad_free_speech
cd953ebf
vad : refactor whiser_vad_params API
f42e6e47
vad : simplify whisper_vad_timestamps_from_probs()
4ff858ba
vad : refactor whisper_vad_timestamps_from_probs to use C++
13a75177
ggerganov
force pushed
from
66bc585e
to
13a75177
1 year ago
vad : make whisper_vad_timestamps oblique in API
3bcc44c2
vad : rename whisper_vad_speech to whisper_vad_probs
5543c80c
vad : move whisper_vad_segment to whisper.cpp
8b6f19cf
vad : make segments vector a std::vector
7625ba16
vad : use std::vector for segments in whisper_vad_timestamps_from_probs
b0b2f9b4
vad : rename pcmf32 parameters to samples [no ci]
20fe0b35
vad : remove n_segments from struct whisper_vad_timestamps
f2123105
vad : rename whisper_vad_timestamps to whisper_vad_segments [no ci]
4c7fe00c
vad : remove whisper_vad_probs struct [no ci]
dc541f93
vad : remove whisper_vad_state struct
163ad538
vad : remove window_size_samples from VAD params
810981f4
vad : clarify VAD CLI options [no ci]
050038ca
docs : add VAD section to README.md [no ci]
3cff6587
squash! docs : add VAD section to README.md [no ci]
acc8747d
vad : minor rename
7aac6eca
ggerganov
approved these changes
on 2025-05-12
squash! docs : add VAD section to README.md [no ci]
41c20100
vad : fix cli option names [no ci]
67f0fd40
danbev
merged
e41bc5c6
into master 1 year ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub