af_whisper.c File Reference

#include <stdio.h>
#include <stdint.h>
#include <stdlib.h>
#include <whisper.h>
#include "libavutil/avutil.h"
#include "libavutil/opt.h"
#include "libavutil/channel_layout.h"
#include "libavutil/samplefmt.h"
#include "libavfilter/avfilter.h"
#include "libavfilter/audio.h"
#include "libavutil/mem.h"
#include "libavutil/avstring.h"
#include "libavutil/internal.h"
#include "libavformat/avio.h"
#include "libavutil/thread.h"
#include "formats.h"

Go to the source code of this file.

Data Structures
struct	WhisperContext

Macros
#define	OFFSET(x)   offsetof(WhisperContext, x)
#define	FLAGS   AV_OPT_FLAG_AUDIO_PARAM | AV_OPT_FLAG_FILTERING_PARAM
#define	HOURS   3600000000

Functions
static void	cb_log (enum ggml_log_level level, const char *text, void *user_data)
static int	init (AVFilterContext *ctx)
static void	uninit (AVFilterContext *ctx)
static void	run_transcription (AVFilterContext *ctx, AVFrame *frame, int samples)
static int	filter_frame (AVFilterLink *inlink, AVFrame *frame)
static int	push_last_frame (AVFilterLink *outlink)
static int	activate (AVFilterContext *ctx)
static int	query_formats (const AVFilterContext *ctx, AVFilterFormatsConfig **cfg_in, AVFilterFormatsConfig **cfg_out)

Variables
static const AVOption	whisper_options []
static const AVClass	whisper_class
const FFFilter	ff_af_whisper

Macro Definition Documentation

OFFSET

#define OFFSET

(

x

)

offsetof(WhisperContext, x)

Definition at line 456 of file af_whisper.c.

FLAGS

#define FLAGS   AV_OPT_FLAG_AUDIO_PARAM | AV_OPT_FLAG_FILTERING_PARAM

Definition at line 457 of file af_whisper.c.

HOURS

#define HOURS   3600000000

Definition at line 458 of file af_whisper.c.

Function Documentation

cb_log()

static void cb_log ( enum ggml_log_level  level, const char *  text, void *  user_data  )

static

Definition at line 76 of file af_whisper.c.

Referenced by init().

init()

static int init ( AVFilterContext *  ctx )

static

Definition at line 91 of file af_whisper.c.

uninit()

static void uninit ( AVFilterContext *  ctx )

static

Definition at line 175 of file af_whisper.c.

run_transcription()

static void run_transcription ( AVFilterContext *  ctx, AVFrame *  frame, int  samples  )

static

Definition at line 201 of file af_whisper.c.

Referenced by filter_frame(), and push_last_frame().

filter_frame()

static int filter_frame ( AVFilterLink *  inlink, AVFrame *  frame  )

static

Definition at line 315 of file af_whisper.c.

Referenced by activate().

push_last_frame()

static int push_last_frame ( AVFilterLink *  outlink )

static

Definition at line 374 of file af_whisper.c.

Referenced by activate().

activate()

static int activate ( AVFilterContext *  ctx )

static

Definition at line 400 of file af_whisper.c.

query_formats()

static int query_formats ( const AVFilterContext *  ctx, AVFilterFormatsConfig **  cfg_in, AVFilterFormatsConfig **  cfg_out  )

static

Definition at line 436 of file af_whisper.c.

Variable Documentation

whisper_options

const AVOption whisper_options[]

static

Initial value:

= {
    { "model", "Path to the whisper.cpp model file", OFFSET(model_path), AV_OPT_TYPE_STRING,.flags = FLAGS },
    { "language", "Language for transcription ('auto' for auto-detect)", OFFSET(language_str), AV_OPT_TYPE_STRING, {.str = "auto"}, .flags = FLAGS },
    { "translate", "Translate from source language to English", OFFSET(translate), AV_OPT_TYPE_BOOL, {.i64 = 0}, 0, 1, .flags = FLAGS },
    { "queue", "Audio queue size", OFFSET(queue), AV_OPT_TYPE_DURATION, {.i64 = 3000000}, 20000, HOURS, .flags = FLAGS },
    { "use_gpu", "Use GPU for processing", OFFSET(use_gpu), AV_OPT_TYPE_BOOL, {.i64 = 1}, 0, 1, .flags = FLAGS },
    { "gpu_device", "GPU device to use", OFFSET(gpu_device), AV_OPT_TYPE_INT, {.i64 = 0}, 0, INT_MAX, .flags = FLAGS },
    { "destination", "Output destination", OFFSET(destination), AV_OPT_TYPE_STRING, {.str = ""}, .flags = FLAGS },
    { "format", "Output format (text|srt|json)", OFFSET(format), AV_OPT_TYPE_STRING, {.str = "text"},.flags = FLAGS },
    { "max_len", "Max segment length in characters", OFFSET(max_len), AV_OPT_TYPE_INT, {.i64 = 0}, 0, INT_MAX, .flags = FLAGS },
    { "vad_model", "Path to the VAD model file", OFFSET(vad_model_path), AV_OPT_TYPE_STRING,.flags = FLAGS },
    { "vad_threshold", "VAD threshold", OFFSET(vad_threshold), AV_OPT_TYPE_FLOAT, {.dbl = 0.5}, 0.0, 1.0, .flags = FLAGS },
    { "vad_min_speech_duration", "Minimum speech duration for VAD", OFFSET(vad_min_speech_duration), AV_OPT_TYPE_DURATION, {.i64 = 100000}, 20000, HOURS, .flags = FLAGS },
    { "vad_min_silence_duration", "Minimum silence duration for VAD", OFFSET(vad_min_silence_duration), AV_OPT_TYPE_DURATION, {.i64 = 500000}, 0, HOURS, .flags = FLAGS },
    { NULL }
}

Definition at line 460 of file af_whisper.c.

whisper_class

const AVClass whisper_class

static

Initial value:

= {
    .class_name = "whisper",
    .item_name = av_default_item_name,
    .option = whisper_options,
    .version = LIBAVUTIL_VERSION_INT,
}

Definition at line 477 of file af_whisper.c.

ff_af_whisper

const FFFilter ff_af_whisper

Initial value:

= {
    .p.name = "whisper",
    .p.description = NULL_IF_CONFIG_SMALL("Transcribe audio using whisper.cpp."),
    .p.priv_class = &whisper_class,
    .p.flags = AVFILTER_FLAG_METADATA_ONLY,
    .init = init,
    .uninit = uninit,
    .activate = activate,
    .priv_size = sizeof(WhisperContext),
    FILTER_INPUTS(ff_audio_default_filterpad),
    FILTER_OUTPUTS(ff_audio_default_filterpad),
    FILTER_QUERY_FUNC2(query_formats),
}

Definition at line 484 of file af_whisper.c.