OrgACRN
/
sof
mirror of https://github.com/thesofproject/sof.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

tools: plugin: README: Add instructions for testing noise suppression 

Fix the stale instruction for install and add the steps for setting up
OpenVino/OpenCV to test the noise suppression model from the open model
zoo repository.

Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
This commit is contained in:
Ranjani Sridharan 2024-02-08 16:48:41 -08:00 committed by Kai Vehmanen
parent e9352299db
commit 48462a47bd
1 changed files with 26 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
tools/plugin/README.md
View File

@ -34,7 +34,7 @@ cmake --build build-plugin/
then (use default ALSA prefix atm)
```
sudo make install <ALSA prefix>
sudo cmake --install build-plugin/
Make sure to set the LD_LIBRARY_PATH to include directory where the SOF modules are installed
Ex: "~/work/sof/sof/build_plugin/sof_ep/install/lib:~/work/sof/sof/build_plugin/modules/"
@ -76,6 +76,31 @@ Mixer settings can be adjusted for bdw-nocodec by (Not functional yet)
alsamixer -Dsof:tgl-nocodec:1
```
# Instructions for testing OpenVino noise suppression model with the SOF plugin:
1. Fetch the model from the Open Model zoo repository ex: noise-suppression-poconetlike-0001.xml
https://docs.openvino.ai/archive/2023.0/omz_demos.html#build-the-demo-applications-on-linux
2. Source OpenVino environment and get OpenCV
https://www.intel.com/content/www/us/en/developer/tools/openvino-toolkit-download.html
3. Worth building and running the demo noise suppression application in the open model zoo
repository to make sure OpenVino and OpenCV are configured properly.
4. Set environment variable NOISE_SUPPRESSION_MODEL_NAME to point to the model file
```
export NOISE_SUPPRESSION_MODEL_NAME="~/open_model_zoo/demos/noise_suppression_demo/cpp/intel/noise-suppression-poconetlike-0001/FP16/noise-suppression-poconetlike-0001.xml"
```
5. Enable noise suppression by setting NOISE_SUPPRESSION=true in the tplg-targets for the sof-plugin topology
6. Start capture using the following command. This uses the 16K capture from the DMIC from
PCM hw device 0,7. Currently, only 16K capture is supported but in the future this will be expanded
to work with 48K capture.
```
arecord -Dsof:plugin:1:0:7:16k2c16b -f dat -r 16000 --period-size=2048 file_ns_16k.wa
```
#TODO Items (and T-shirt size) for single pipeline E2E audio
* IPC4 support in tplg parser (M)
* IPC4 support in plugin (pipe/ipc4.c) (M)