Skip to content

Fixed VAD to work when using whisper_full_with_state #3423

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
Open

Fixed VAD to work when using whisper_full_with_state #3423

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b91fe0f
Fixed VAD to work when using whisper_full_with_state
mdestagnol Sep 17, 2025
dc1e51f
fix
mdestagnol Sep 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
81 changes: 38 additions & 43 deletions src/whisper.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6652,8 +6652,8 @@ static bool whisper_vad(

if (vad_segments->data.size() > 0) {
state->has_vad_segments = true;
ctx->state->vad_segments.clear();
ctx->state->vad_segments.reserve(vad_segments->data.size());
state->vad_segments.clear();
state->vad_segments.reserve(vad_segments->data.size());

// Initialize the time mapping table
state->vad_mapping_table.clear();
Expand Down Expand Up @@ -6749,7 +6749,7 @@ static bool whisper_vad(

WHISPER_LOG_INFO("%s: vad_segment_info: orig_start: %.2f, orig_end: %.2f, vad_start: %.2f, vad_end: %.2f\n",
__func__, segment.orig_start/100.0, segment.orig_end/100.0, segment.vad_start/100.0, segment.vad_end/100.0);
ctx->state->vad_segments.push_back(segment);
state->vad_segments.push_back(segment);

// Copy this speech segment
memcpy(filtered_samples.data() + offset, samples + segment_start_samples, segment_length * sizeof(float));
Expand Down Expand Up @@ -6802,19 +6802,41 @@ static bool whisper_vad(
}

int whisper_full_with_state(
struct whisper_context * ctx,
struct whisper_state * state,
struct whisper_full_params params,
const float * samples,
int n_samples) {
struct whisper_context *ctx,
struct whisper_state *state,
struct whisper_full_params params,
const float *samples,
int n_samples)
{

std::vector<float> vad_samples;
if (params.vad)
{
WHISPER_LOG_INFO("%s: VAD is enabled, processing speech segments only\n", __func__);
if (!whisper_vad(ctx, state, params, samples, n_samples, vad_samples))
{
WHISPER_LOG_ERROR("%s: failed to compute VAD\n", __func__);
return -1;
}
if (vad_samples.empty())
{
state->result_all.clear();
return 0;
}
samples = vad_samples.data();
n_samples = vad_samples.size();
}

// clear old results
auto & result_all = state->result_all;
auto &result_all = state->result_all;

result_all.clear();

if (n_samples > 0) {
if (n_samples > 0)
{
// compute log mel spectrogram
if (whisper_pcm_to_mel_with_state(ctx, state, samples, n_samples, params.n_threads) != 0) {
if (whisper_pcm_to_mel_with_state(ctx, state, samples, n_samples, params.n_threads) != 0)
{
WHISPER_LOG_ERROR("%s: failed to compute log mel spectrogram\n", __func__);
return -2;
}
Expand Down Expand Up @@ -7720,25 +7742,11 @@ int whisper_full_with_state(
}

int whisper_full(
struct whisper_context * ctx,
struct whisper_full_params params,
const float * samples,
int n_samples) {

std::vector<float> vad_samples;
if (params.vad) {
WHISPER_LOG_INFO("%s: VAD is enabled, processing speech segments only\n", __func__);
if (!whisper_vad(ctx, ctx->state, params, samples, n_samples, vad_samples)) {
WHISPER_LOG_ERROR("%s: failed to compute VAD\n", __func__);
return -1;
}
if (vad_samples.empty()) {
ctx->state->result_all.clear();
return 0;
}
samples = vad_samples.data();
n_samples = vad_samples.size();
}
struct whisper_context *ctx,
struct whisper_full_params params,
const float *samples,
int n_samples)
{
return whisper_full_with_state(ctx, ctx->state, params, samples, n_samples);
}

Expand All @@ -7753,19 +7761,6 @@ int whisper_full_parallel(
return whisper_full(ctx, params, samples, n_samples);
}

std::vector<float> vad_samples;
if (params.vad) {
WHISPER_LOG_INFO("%s: VAD is enabled, processing speech segments only\n", __func__);
if (!whisper_vad(ctx, ctx->state, params, samples, n_samples, vad_samples)) {
WHISPER_LOG_ERROR("%s: failed to compute VAD\n", __func__);
return -1;
}
if (vad_samples.empty()) {
return 0;
}
samples = vad_samples.data();
n_samples = vad_samples.size();
}
int ret = 0;

// prepare separate states for each thread
Expand Down