Connecting to CMU Sphinx using PHP

问题

I have been looking into Speech Recognition and ways in which this can be implemented into a website. I have found many examples on using it with Python and even one with Node.js but I want to be able to use PHP with this.

Is there any way I can access CMUSphinx on a Linux server using PHP to process my inputs?

Thanks

回答1:

Can be done but to use asterisks as the audio capture and processing engine. See http://www.voip-info.org/wiki/view/Sphinx

Example code below after your server has been configured

    function sphinx($filename='', $timeout=3000, $service_port = 1069, $address = '127.0.0.1'){

        /* if a recording has not been passed in we create one */
        if ($filename=="") {
            $filename = "/var/lib/asterisk/sounds/sphinx_".$this->request['agi_uniqueid'];
            $extension = "wav";
            $this->stream_file('beep', 3000, 5);
            $this->record_file($filename, $extension, '0',$timeout);
            $filename=$filename.'.'.$extension;
        }   

        /* Create a TCP/IP socket. */
        $socket = socket_create(AF_INET, SOCK_STREAM, SOL_TCP);
        if ($socket < 0) {
            return false;
        }

        $result = socket_connect($socket, $address, $service_port);
        if ($result < 0) {
           return false;
        }

        //open the file and read in data
        $handle = fopen($filename, "rb");
        $data = fread($handle, filesize($filename));

        socket_write($socket, filesize($filename)."\n");
        socket_write($socket, $data);

        $response = socket_read($socket, 2048);

        socket_close($socket);

        unlink($filename);
        return $response;
   }

Another thought after looking at the website is that sphinx 4 allows web service access to the recognition processing daemon ie: run sphinx as a daemon (its java!) then you can do socket opens as above to feed a .wav into it directly basically using a modification of the code above so instead of calling the asterisks server to retrieve then record the audio you would use something else perhaps html5 etc to record the audio.

Another thing to consider is that chrome and html5 has built in speech recognition

回答2:

The architecture of such system depends on the audio type you want to process. If audio is long you can just store it into temporary file and invoke pocketsphinx_continuous as an external tool to process it:

http://php.net/manual/en/function.shell-exec.php

You call pocketsphinx_continuous -infile file.wav > decode-result.txt and that give you a result to display. The problem with this approach is that decoding initialization takes time so you will not be able to use that approach for short files.

If you want to process short samples or you want to process audio in streaming mode, you need some kind of server to load models and wait for requests. There are different variants on how to implement it from simple hand-made server listening on TCP port with simple protocol and accepting data to more complex solutions like http://unimrcp.org

来源：https://stackoverflow.com/questions/25666760/connecting-to-cmu-sphinx-using-php

标签

php

voice-recognition

cmusphinx