Commit a64363b9 authored by Abdel HEBA's avatar Abdel HEBA
Browse files
parents 2b28a08f 63898e54
......@@ -14,9 +14,9 @@ Using this project, you will be able to run an offline Automatic Speech Recognit
Attention
--------
The ASR server that will be setup here require some kaldi model, In the docker image that I will detail below, there is no kaldi model included.
The ASR server that will be setup here require kaldi model, In the docker image that I will detail below, there is no kaldi model included.
You must have this model on your machine. You must also check that the model have this specific files :
You must have this model on your machine. You must also check that the model have the specific files bellow :
- final.alimdl
- final.mat
- final.mdl
......@@ -31,52 +31,34 @@ You must have this model on your machine. You must also check that the model hav
Install docker
---------
Please, refer to [doc docker](https://docs.docker.com/engine/installation).
Please, refer to [docker doc](https://docs.docker.com/engine/installation).
Get the image
---------
Currently, the image docker is about (4GB) and have debian8 OS, and not yet pulled on DockerHub.
Currently, the image docker is about (4GB) and based on debian8, the image docker has not yet pulled on DockerHub.
You need to build your own image:
`docker build -t linagora/stt-offline .`
Deploy Achritecture
------------
You need to specify both of your Kaldi and STT model directories:
```
./deploy-offline-decoding.sh <KALDI_PATH> <STT_Model_PATH>
```
the `deploy-offline-decoding.sh` script generate:
- wavs directory: you need to put all wavs that you need to transcribe there
- trans directory: you will find all transcripts there
- scripts: contain all scripts for decoding
- systems: will contain the STT Model and decoding directory for each wav
```
├── wavs
├── trans
├── deploy_offline_decoding.sh
├── Readme.md
├── scripts
│   ├── cmd.sh
│   ├── conf
│   │   └── mfcc.conf
│   ├── decode.sh
│   ├── path.sh
│   ├── steps -> <KALDI_PATH>/egs/wsj/s5/steps/
│   └── utils -> <KALDI_PATH>/egs/wsj/s5/utils/
├── systems
│   └── SYS1=LEX001+LM001+AM001
│   └── tri3 -> <STT_Model_PATH>
└── tools
├── kaldi -> <KALDI_PATH>
└── LIUM_SpkDiarization-8.4.1.jar
docker build -t linagora/stt-offline .
```
How to use
----------
`start_docker.sh` allow to build and create the container assuming that your kaldi model is located at `<Path_model>`
```
./start_docker.sh <Path_model> <Port>
```
The `<Port>` param publish a container's port to the host, you should use POST method to send wav file to the server for transcription.
Run Example
----------
to be described
\ No newline at end of file
Simple call using curl:
```
curl -F "wav_file=@<wav_path>" http://<IP:PORT_service>/upload > <output_trans>
```
The attribut `wav_file` is needed to submit the wav file to the server using POST Method
Client script is available and allow to connect to the server located at `http://localhost:<Port>/upload`
```
./client/client <wav_path> <IP_server>:<POST> <Output>
```
\ No newline at end of file
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment