No tienes acceso a esta clase

隆Contin煤a aprendiendo! 脷nete y comienza a potenciar tu carrera

Watson Speech to Text

15/25
Recursos

Aportes 22

Preguntas 7

Ordenar por:

Los aportes, preguntas y respuestas son vitales para aprender en comunidad. Reg铆strate o inicia sesi贸n para participar.

El servicio Speech to Text convierte la voz humana en palabras escritas. Se puede utilizar en cualquier lugar donde se necesite llenar el hueco entre la palabra hablada y la forma escrita, incluido el control de voz de los sistemas incorporados, la transcripci贸n de reuniones y teleconferencias y el dictado de correo electr贸nico y notas.

El modelo de idioma espa帽ol Colombia es: model=es-CO_BroadbandModel

la verdad no me funciono coomo esperaba, envie dos audios.

  1. hola soy julian orozco y soy de 鈥渃iudad鈥
    鈥渃onfidence鈥: 0.36,
    鈥渢ranscript鈥: "when I started my number is called auto school showed up there you go "
  2. estados unidos de norte america
    "confidence": 0.5,
    鈥渢ranscript鈥: "espousal near the North America "
    nos se si hay que configurar el idioma de entrada o que paso aca.

Qu茅 bonito es cuando s铆 funciona jaja
![](

Un peque帽o comentario, se pone la url justo antes de v1, pues por ejemplo a mi no me aparece 鈥渁pi鈥 al final de mi url:

curl -X POST -u "apikey:SPkIfPpktELAu5mS0fR2D6eQz5fvGFUUdbxGzzmWz2wz" \
 --header "Content-Type: audio/flac" --data-binary @audio-file2.flac \
  "https://api.us-south.speech-to-text.watson.cloud.ibm.com/instances/219b75ae-b66c-4993-b2d6-06025c1813a4/v1/recognize?model=es-MX_BroadbandModel&keywords=%22isaac%22&keywords_threshold=0.5"

驴Las malas palabras estan incluidas en Watson Speech to Text? no lo encontre en el docs https://cloud.ibm.com/docs/services/speech-to-text?topic=speech-to-text-gettingStarted

Con los audios de la clase funciona bien pero al grabar un audio con el celular y enviarlo ya no funciona, dice error 400, en el reproductor que uso (VCL) se reproduce bien, la solucion fue tomar el archivo original y pasarlo por un conversor a flac y en este caso si funciono

Me funciono igual.
鈥渢ranscript鈥: "cu茅llar isaac gracias por tomar este curso s铆gueme mi redes en instagram estoy como el buen isi y en lo dem谩s como isaac carranza "

Excelente clase

驴Alguien me podr铆a ayudar? me sale lo siguiente:
鈥渒eywords鈥 no se reconoce como un comando interno o externo,
programa o archivo por lotes ejecutable.
鈥渒eywords_threshold鈥 no se reconoce como un comando interno o externo,
programa o archivo por lotes ejecutable.

Solo recibe audios .flac? Como se configura que idioma es el que debe transcribir?

Muy poderosa la herramienta de Speech to text.

Instagram @elbuenizzy

Hola, alguien me podr铆a ayudar no me sale he hecho de todo. Me sale un mensaje muy largo solo colocar茅 una parte:
-F, --form <name=content> Specify HTTP multipart POST data
鈥揻orm-string <name=string> Specify HTTP multipart POST data
鈥揻tp-account <data> Account data string
鈥揻tp-alternative-to-user <command> String to replace USER [name]
鈥揻tp-create-dirs Create the remote dirs if not present
鈥揻tp-method <method> Control CWD usage
鈥揻tp-pasv Use PASV/EPSV instead of PORT
-P, --ftp-port <address> Use PORT instead of PASV
鈥揻tp-pret Send PRET before PASV
鈥揻tp-skip-pasv-ip Skip the IP address for PASV
鈥揻tp-ssl-ccc Send CCC after authenticating
鈥揻tp-ssl-ccc-mode <active/passive> Set CCC mode
鈥揻tp-ssl-control Require SSL/TLS for FTP login, clear for transfer
-G, --get Put the post data in the URL and use GET
-g, --globoff Disable URL sequences and ranges using {} and []
-I, --head Show document info only
-H, --header header/@file Pass custom header(s) to server
-h, --help This help text
鈥揾ostpubmd5 <md5> Acceptable MD5 hash of the host public key
-0, --http1.0 Use HTTP 1.0
鈥揾ttp1.1 Use HTTP 1.1
鈥揾ttp2 Use HTTP 2
鈥揾ttp2-prior-knowledge Use HTTP 2 without HTTP/1.1 Upgrade
鈥搃gnore-content-length Ignore the size of the remote resource
-i, --include Include protocol response headers in the output
-k, --insecure Allow insecure server connections when using SSL
鈥搃nterface <name> Use network INTERFACE (or address)
-4, --ipv4 Resolve names to IPv4 addresses
-6, --ipv6 Resolve names to IPv6 addresses
-j, --junk-session-cookies Ignore session cookies read from file
鈥搆eepalive-time <seconds> Interval time for keepalive probes
鈥搆ey <key> Private key file name
鈥搆ey-type <type> Private key file type (DER/PEM/ENG)
鈥搆rb <level> Enable Kerberos with security <level>
鈥搇ibcurl <file> Dump libcurl equivalent code of this command line
鈥搇imit-rate <speed> Limit transfer speed to RATE
-l, --list-only List only mode
鈥搇ocal-port <num/range> Force use of RANGE for local port numbers
-L, --location Follow redirects
鈥搇ocation-trusted Like --location, and send auth to other hosts
鈥搇ogin-options <options> Server login options
鈥搈ail-auth <address> Originator address of the original email
鈥搈ail-from <address> Mail from this address
鈥搈ail-rcpt <address> Mail from this address
-M, --manual Display the full manual
鈥搈ax-filesize <bytes> Maximum file size to download
鈥搈ax-redirs <num> Maximum number of redirects allowed
-m, --max-time <time> Maximum time allowed for the transfer
鈥搈etalink Process given URLs as metalink XML file
鈥搉egotiate Use HTTP Negotiate (SPNEGO) authentication
-n, --netrc Must read .netrc for user name and password
鈥搉etrc-file <filename> Specify FILE for netrc
鈥搉etrc-optional Use either .netrc or URL
-:, --next Make next URL use its separate set of options
鈥搉o-alpn Disable the ALPN TLS extension
-N, --no-buffer Disable buffering of the output stream
鈥搉o-keepalive Disable TCP keepalive on the connection
鈥搉o-npn Disable the NPN TLS extension
鈥搉o-sessionid Disable SSL session-ID reusing
鈥搉oproxy <no-proxy-list> List of hosts which do not use proxy
鈥搉tlm Use HTTP NTLM authentication
鈥搉tlm-wb Use HTTP NTLM authentication with winbind
鈥搊auth2-bearer <token> OAuth 2 Bearer Token
-o, --output <file> Write to file instead of stdout
鈥損ass <phrase> Pass phrase for the private key
鈥損ath-as-is Do not squash 鈥 sequences in URL path
鈥損innedpubkey <hashes> FILE/HASHES Public key to verify peer against
鈥損ost301 Do not switch to GET after following a 301
鈥損ost302 Do not switch to GET after following a 302
鈥損ost303 Do not switch to GET after following a 303
鈥損reproxy [protocol://]host[:port] Use this proxy first
-#, --progress-bar Display transfer progress as a bar
鈥損roto <protocols> Enable/disable PROTOCOLS
鈥損roto-default <protocol> Use PROTOCOL for any URL missing a scheme
鈥損roto-redir <protocols> Enable/disable PROTOCOLS on redirect
-x, --proxy [protocol://]host[:port] Use this proxy
鈥損roxy-anyauth Pick any proxy authentication method
鈥損roxy-basic Use Basic authentication on the proxy
鈥損roxy-cacert <file> CA certificate to verify peer against for proxy
鈥損roxy-capath <dir> CA directory to verify peer against for proxy
鈥損roxy-cert <cert[:passwd]> Set client certificate for proxy
鈥損roxy-cert-type <type> Client certificate type for HTTS proxy
鈥損roxy-ciphers <list> SSL ciphers to use for proxy
鈥損roxy-crlfile <file> Set a CRL list for proxy
鈥損roxy-digest Use Digest authentication on the proxy
鈥損roxy-header header/@file Pass custom header(s) to proxy
鈥損roxy-insecure Do HTTPS proxy connections without verifying the proxy
鈥損roxy-key <key> Private key for HTTPS proxy
鈥損roxy-key-type <type> Private key file type for proxy
鈥損roxy-negotiate Use HTTP Negotiate (SPNEGO) authentication on the proxy
鈥損roxy-ntlm Use NTLM authentication on the proxy
鈥損roxy-pass <phrase> Pass phrase for the private key for HTTPS proxy
鈥損roxy-service-name <name> SPNEGO proxy service name
鈥損roxy-ssl-allow-beast Allow security flaw for interop for HTTPS proxy
鈥損roxy-tlsauthtype <type> TLS authentication type for HTTPS proxy
鈥損roxy-tlspassword <string> TLS password for HTTPS proxy
鈥損roxy-tlsuser <name> TLS username for HTTPS proxy
鈥損roxy-tlsv1 Use TLSv1 for HTTPS proxy
-U, --proxy-user user:password Proxy user and password
鈥損roxy1.0 <host[:port]> Use HTTP/1.0 proxy on given port
-p, --proxytunnel Operate through a HTTP proxy tunnel (using CONNECT)
鈥損ubkey <key> SSH Public key file name
-Q, --quote Send command(s) to server before transfer
鈥搑andom-file <file> File for reading random data from
-r, --range <range> Retrieve only the bytes within RANGE
鈥搑aw Do HTTP 鈥渞aw鈥; no transfer decoding
-e, --referer <URL> Referrer URL
-J, --remote-header-name Use the header-provided filename
-O, --remote-name Write output to a file named as the remote file
鈥搑emote-name-all Use the remote file name for all URLs
-R, --remote-time Set the remote file鈥檚 time on the local output
-X, --request <command> Specify request command to use
鈥搑equest-target Specify the target for this request
鈥搑esolve host:port:address Resolve the host+port to this address
鈥搑etry <num> Retry request if transient problems occur
鈥搑etry-connrefused Retry on connection refused (use with --retry)
鈥搑etry-delay <seconds> Wait time between retries
鈥搑etry-max-time <seconds> Retry only within this period
鈥搒asl-ir Enable initial response in SASL authentication
鈥搒ervice-name <name> SPNEGO service name
-S, --show-error Show error even when -s is used
-s, --silent Silent mode
鈥搒ocks4 <host[:port]> SOCKS4 proxy on given host + port
鈥搒ocks4a <host[:port]> SOCKS4a proxy on given host + port
鈥搒ocks5 <host[:port]> SOCKS5 proxy on given host + port
鈥搒ocks5-basic Enable username/password auth for SOCKS5 proxies
鈥搒ocks5-gssapi Enable GSS-API auth for SOCKS5 proxies
鈥搒ocks5-gssapi-nec Compatibility with NEC SOCKS5 server
鈥搒ocks5-gssapi-service <name> SOCKS5 proxy service name for GSS-API
鈥搒ocks5-hostname <host[:port]> SOCKS5 proxy, pass host name to proxy
-Y, --speed-limit <speed> Stop transfers slower than this
-y, --speed-time <seconds> Trigger 鈥榮peed-limit鈥 abort after this time
鈥搒sl Try SSL/TLS
鈥搒sl-allow-beast Allow security flaw to improve interop
鈥搒sl-no-revoke Disable cert revocation checks (WinSSL)
鈥搒sl-reqd Require SSL/TLS
-2, --sslv2 Use SSLv2
-3, --sslv3 Use SSLv3
鈥搒tderr Where to redirect stderr
鈥搒uppress-connect-headers Suppress proxy CONNECT response headers
鈥搕cp-fastopen Use TCP Fast Open
鈥搕cp-nodelay Use the TCP_NODELAY option
-t, --telnet-option <opt=val> Set telnet option
鈥搕ftp-blksize <value> Set TFTP BLKSIZE option
鈥搕ftp-no-options Do not send any TFTP options
-z, --time-cond <time> Transfer based on a time condition
鈥搕ls-max <VERSION> Use TLSv1.0 or greater
鈥搕lsauthtype <type> TLS authentication type
鈥搕lspassword TLS password
鈥搕lsuser <name> TLS user name
-1, --tlsv1 Use TLSv1.0 or greater
鈥搕lsv1.0 Use TLSv1.0
鈥搕lsv1.1 Use TLSv1.1
鈥搕lsv1.2 Use TLSv1.2
鈥搕lsv1.3 Use TLSv1.3
鈥搕r-encoding Request compressed transfer encoding
鈥搕race <file> Write a debug trace to FILE
鈥搕race-ascii <file> Like --trace, but without hex output
鈥搕race-time Add time stamps to trace/verbose output
鈥搖nix-socket <path> Connect through this Unix domain socket
-T, --upload-file <file> Transfer local FILE to destination
鈥搖rl <url> URL to work with
-B, --use-ascii Use ASCII/text transfer
-u, --user user:password Server user and password
-A, --user-agent <name> Send User-Agent <name> to server
-v, --verbose Make the operation more talkative
-V, --version Show version number and quit
-w, --write-out <format> Use output FORMAT after completion
鈥搙attr Store metadata in extended file attributes

Super interesante 馃槂

Seria interesante ver c贸mo agregar la inteligencia a los textos, por ejemplo poner el keyword nombre y que el me devuelva el nombre. Asi como en el ML UNL

para lo que utilizan windows (como yo):

curl -X POST -u "apikey:{apikey}" --header "Content-Type: audio/flac" --data-binary @audio-file2.flac "{url}/v1/recognize?model=es-MX_BroadbandModel&keywords=%22isaac%22&keywords_threshold=0.5"

Aporte. Lo prob茅 con formato .ogg y funciona muy bien.

Saludos!

Buen servicio!

Que bueno cuando todo te funciona.

Funcion贸 perfecto en los dos casos de ejemplo, realizados por el profesor. Al parecer el servicio funciona con archivos de audio guardados con la extensi贸n .flac.

Lo prob茅 con un archivo .wav y funcion贸 perfecto!