File size: 3,712 Bytes
71e354e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 | /**************************************************************************************************
*
* Copyright (c) 2019-2026 Axera Semiconductor (Ningbo) Co., Ltd. All Rights Reserved.
*
* This source file is the property of Axera Semiconductor (Ningbo) Co., Ltd. and
* may not be copied or distributed in any isomorphic form without the prior
* written consent of Axera Semiconductor (Ningbo) Co., Ltd.
*
**************************************************************************************************/
#ifndef _AX_TTS_API_H_
#define _AX_TTS_API_H_
#ifdef __cplusplus
extern "C" {
#endif
#define AX_TTS_API __attribute__((visibility("default")))
#define AX_TTS_MAX_STR_LEN 256
// Supported TTS models
enum AX_TTS_TYPE_E {
AX_KOKORO = 0,
AX_MELOTTS
};
// TTS Init config
typedef struct {
int max_seq_len;
char model_path[AX_TTS_MAX_STR_LEN];
char espeak_data_path[AX_TTS_MAX_STR_LEN];
char language[AX_TTS_MAX_STR_LEN];
char jieba_dict_path[AX_TTS_MAX_STR_LEN];
} AX_TTS_INIT_CONFIG;
// TTS Run config
typedef struct {
float speed;
float fade_out;
int sample_rate;
char voice[AX_TTS_MAX_STR_LEN];
char language[AX_TTS_MAX_STR_LEN];
} AX_TTS_RUN_CONFIG;
// Speech audio
typedef struct {
int sample_rate;
int num_samples;
int channels;
float data[];
} AX_TTS_AUDIO;
/**
* @brief Opaque handle type for TTS context
*
* This handle encapsulates all internal state of the TTS system.
* The actual implementation is hidden from C callers to maintain ABI stability.
*/
typedef void* AX_TTS_HANDLE;
/**
* @brief Initialize the TTS system with specific configuration
*
* Creates and initializes a new TTS context with the specified
* model type, model path, and language. This function loads the appropriate
* models, configures the generator, and prepares it for speech generation.
*
* @param model_type Type of model to use
* @param model_path Directory path where model files are stored
* Model files are expected to be in the format: *.axmodel
*
* @return AX_TTS_HANDLE Opaque handle to the initialized context,
* or NULL if initialization fails
*
* @note The caller is responsible for calling AX_TTS_Uninit() to free
* resources when the handle is no longer needed.
* @example
* // Initialize recognition with whisper tiny model
* AX_TTS_HANDLE handle = AX_TTS_Init(AX_KOKORO, "./models-ax650/");
*
*/
AX_TTS_API AX_TTS_HANDLE AX_TTS_Init(AX_TTS_TYPE_E tts_type, AX_TTS_INIT_CONFIG* init_config);
/**
* @brief Deinitialize and release TTS resources
*
* Cleans up all resources associated with the context, including
* unloading models, freeing memory, and releasing hardware resources.
*
* @param handle context handle obtained from AX_TTS_Init()
*
* @warning After calling this function, the handle becomes invalid and
* should not be used in any subsequent API calls.
*/
AX_TTS_API void AX_TTS_Uninit(AX_TTS_HANDLE handle);
/**
* @brief Perform speech generation and return dynamically allocated struct
*
* @param handle context handle
* @param text Text input to generate speech
* @param tts_config Config of generation
* @param audio Pointer to receive the allocated audio
*
* @return int Status code (0 = success, <0 = error)
*
* @note The returned audio is allocated with malloc() and must be freed
* by the caller using free() when no longer needed.
*/
AX_TTS_API int AX_TTS_Run(AX_TTS_HANDLE handle,
const char* text,
AX_TTS_RUN_CONFIG* run_config,
AX_TTS_AUDIO** audio);
#ifdef __cplusplus
}
#endif
#endif // _AX_TTS_API_H_
|