feat(mn): update set_det_threshold func

2025-09-15 15:28:44 +08:00 · 2025-04-10 16:23:35 +08:00 · 2025-04-10 16:23:35 +08:00 · 5b5b69630b
commit 5b5b69630b
parent f55acfe05d
19 changed files with 10 additions and 8 deletions
--- a/README.md
+++ b/README.md
@ -23,7 +23,6 @@ The new algorithms will no longer support ESP32 chips.

 News
 ----
-[28/3/2025]: We release **Sound Souce Localization**, an algorithm to locate the sound source.
 [14/2/2025]: We release **ESP-SR V2.0**. [Migration from ESP-SR V1.* to ESP-SR V2.*](https://docs.espressif.com/projects/esp-sr/en/latest/esp32s3/audio_front_end/migration_guide.html)   
 [13/2/2025]: We release **VADNet**, a voice activaty detection model. You can use it to replace the WebRTC VAD and improve the performance.

@ -89,7 +88,7 @@ The following MultiNet models are supported in esp-sr:

 ## Audio Front End

-Espressif Audio Front-End **AFE** integrates AEC (Acoustic Echo Cancellation), VAD (Voice Activity Detection), BSS (Blind Source Separation) and NS (Noise Suppression), NSNET(Deep noise suppression), SSL(Sound Souce Localization) and other functions. It is designed to be used with the ESP-SR library.
+Espressif Audio Front-End **AFE** integrates AEC (Acoustic Echo Cancellation), VAD (Voice Activity Detection), BSS (Blind Source Separation) and NS (Noise Suppression), NSNET(Deep noise suppression) and other functions. It is designed to be used with the ESP-SR library.

 Our two-mic Audio Front-End (AFE) have been qualified as a “Software Audio Front-End Solution” for [Amazon Alexa Built-in devices](https://developer.amazon.com/en-US/alexa/solution-providers/alexa-connect-kit).

--- a/lib/esp32p4/libesp_audio_front_end.a
+++ b/lib/esp32p4/libesp_audio_front_end.a
--- a/lib/esp32p4/libesp_audio_processor.a
+++ b/lib/esp32p4/libesp_audio_processor.a
--- a/lib/esp32p4/libmultinet.a
+++ b/lib/esp32p4/libmultinet.a
--- a/lib/esp32p4/libvadnet.a
+++ b/lib/esp32p4/libvadnet.a
--- a/lib/esp32p4/libwakenet.a
+++ b/lib/esp32p4/libwakenet.a
--- a/lib/esp32s3/libc_speech_features.a
+++ b/lib/esp32s3/libc_speech_features.a
--- a/lib/esp32s3/libdl_lib.a
+++ b/lib/esp32s3/libdl_lib.a
--- a/lib/esp32s3/libesp_audio_front_end.a
+++ b/lib/esp32s3/libesp_audio_front_end.a
--- a/lib/esp32s3/libesp_audio_processor.a
+++ b/lib/esp32s3/libesp_audio_processor.a
--- a/lib/esp32s3/libflite_g2p.a
+++ b/lib/esp32s3/libflite_g2p.a
--- a/lib/esp32s3/libfst.a
+++ b/lib/esp32s3/libfst.a
--- a/lib/esp32s3/libhufzip.a
+++ b/lib/esp32s3/libhufzip.a
--- a/lib/esp32s3/libmultinet.a
+++ b/lib/esp32s3/libmultinet.a
--- a/lib/esp32s3/libnsnet.a
+++ b/lib/esp32s3/libnsnet.a
--- a/lib/esp32s3/libvadnet.a
+++ b/lib/esp32s3/libvadnet.a
--- a/lib/esp32s3/libwakenet.a
+++ b/lib/esp32s3/libwakenet.a
--- a/test_apps/esp-sr/main/test_multinet.cpp
+++ b/test_apps/esp-sr/main/test_multinet.cpp
@ -235,6 +235,7 @@ TEST_CASE("multinet set commands from sdkconfig and detect", "[mn]")
    esp_mn_error_t *error_phrases = NULL;
    esp_mn_commands_update_from_sdkconfig(multinet, model_data);
    multinet->print_active_speech_commands(model_data);
+    multinet->set_det_threshold(model_data, 0.1);

    while (1) {
        if ((chunks + 1)*audio_chunksize <= data_size) {
--- a/test_apps/esp-sr/sdkconfig.ci.p4_afe
+++ b/test_apps/esp-sr/sdkconfig.ci.p4_afe
@ -5,19 +5,21 @@ CONFIG_IDF_TARGET="esp32p4"
 CONFIG_ESPTOOLPY_FLASHMODE_QIO=y
 CONFIG_ESPTOOLPY_FLASHSIZE_16MB=y
 CONFIG_PARTITION_TABLE_CUSTOM=y
+CONFIG_SR_NSN_NSNET2=y
 CONFIG_SR_VADN_VADNET1_MEDIUM=y
 CONFIG_SR_WN_WN9_HILEXIN=y
-CONFIG_SR_NSN_NSNET2=y
-CONFIG_SPIRAM=y
-CONFIG_ESP_TASK_WDT_EN=n
-CONFIG_ESP_TASK_WDT_INIT=n
-CONFIG_ESP_MAIN_TASK_STACK_SIZE=10240
 CONFIG_COMPILER_OPTIMIZATION_PERF=y
+CONFIG_COMPILER_ORPHAN_SECTIONS_PLACE=y
 CONFIG_ESP32P4_REV_MIN_0=y
 CONFIG_SPIRAM=y
 CONFIG_SPIRAM_SPEED_200M=y
 CONFIG_CACHE_L2_CACHE_256KB=y
 CONFIG_CACHE_L2_CACHE_LINE_128B=y
 CONFIG_ESP_SYSTEM_ALLOW_RTC_FAST_MEM_AS_HEAP=n
+CONFIG_ESP_MAIN_TASK_STACK_SIZE=10240
+CONFIG_ESP_TASK_WDT_EN=n
+CONFIG_LOG_COLORS=y
+CONFIG_LWIP_HOOK_IP6_INPUT_NONE=y
 CONFIG_MBEDTLS_CMAC_C=y
+CONFIG_MBEDTLS_ECP_FIXED_POINT_OPTIM=y
 CONFIG_IDF_EXPERIMENTAL_FEATURES=y