Release notes
1.36.1 — 2024-10-24
- Fixed bug in FR date builtin, were
neuf
was not properly recognized.
1.36.0 — 2024-10-15
- Speech recognized so far can be available in H2B Console in case of
Stopped
;
1.35.2 — 2024-09-04
- Spelling grammars with punct are now supported in English, German and Spanish too.
- Date grammar now available for English. In that case, the country code should be provided.
- Date grammar now available for German
- Date grammar now available for Spanish
- A bug was fixed that mangled some words containing the sequence
hum
in the output.
1.34.0 — 2024-08-13
- H2B supports German grammars
- Support for basic builtins
speech/transcribe
,speech/text2num
andspeech/keywords
; - Support for boolean grammar (
speech/boolean
); - Postal Codes (
speech/zipcode
); - Homophones disambiguation for spelling grammars;
- Support for basic builtins
- Normalize city in address grammar result (if found).
1.33.0 — 2024-07-23
- Updated French city database for
speech/postal_address
builtin; - Handles Spanish prefix when spelling "double": "doble";
- the builtins
speech/spelling/zipcode
andspeech/spelling/licence_plate
are now region aware. - Spain postal codes.
1.32.2 — 2024-07-01
- Fixed a bug in the spelling builtin that truncated the input on rare occasions;
- better usage of resources (internal).
1.32.1 — 2024-04-22
- The actual ASR Model name is reported in the result payload and in the developer console.
1.30.1 — 2024-03-19
- New regex engine that can emit partial matches and fix some common confusions according to what the regex expects.
- Support the
Logging-Tag
header as specified in the RFC. - Support for a client side custom session id.
- Refined parameters for early speech detection by the VAD.
1.28.1 — 2024-01-24
- VAD now supports early speech, that is, speech that starts before the RECOGNIZE command is received. Only works if the signal is loud enough.
1.28.0 — 2024-01-03
- Support for double spelled signs: "double A, double 2…" in the builtin spelling
1.27.4 — 2023-11-16
- Safe XML serialization of interpretation.
1.27.2 — 2023-10-27
- Adds 'yes' as a YES answer to the French boolean grammar
1.27.0 — 2023-08-10
- More spelled symbols recognized in
builtin:speech/spelling_*_with_punct
grammars
1.26.11 — 2023-06-13
- Hot fix: single "zéro" was not recognized by builtin spelling
1.26.10 — 2023-05-10
- Fixed channel-id generation in WebSocket front
1.26.9 — 2023-02-22
- Fixed a bug in date builtin, when the month of january is spelled as a number and a full year is given.
- Fixed inappropriate activation of proprietary extension
speech-nomatch-timeout
.
1.26.4 — 2023-02-22
postal_address
has more robust NLU- bug fixes
- auto disconnection of idle clients
- API version embedded in results
- New enhanced VAD
- The
Sensitivity-Level
is now active again.
- The
1.24.0 — 2022-12-13
- Improved
licence_plate
grammar: automatically replaces intial S by F. - Improved
postal_address
grammar:- support for address complements
- support for boroughs (Lyon, Marseille, Paris)
- support for weird postal code enunciation (e.g. "soixante-quinze mille zéro dix-sept")
1.22.0 — 2022-10-24
- New builtin:
speech/spelling/license_plate
1.21.0 — 2022-09-08
- Builtin Date: recognize more expressions like "samedi dans 5 jours", "lundi de la semaine prochaine", "mercredi dans deux semaines", etc…
- Builtin Boolean: give more priority to the strongs, dedicated "yes" word in case of ambiguity like "pas de soucis oui"
- Basic support for Spanish (transcribe, text2num, spelling/*, boolean)
1.20.0 — 2022-08-08
- New builtin grammar:
speech/spelling/letters_with_punct
- Better timer for
speech-nomatch-timeout
1.19.0 — 2022-07-05
- Spelling grammars: fixed confusion of I as Y.
- Fixed handling of apostrophe in the
postal_address
grammar
1.18.0 – 2022-06-13
- New parameter for
speech/spelling/*
:sep
to specify the letter/number separator (ornone
) when noregex
orlength
are given. - Bug fixes:
speech/spelling/mixed_with_punct
now correctly recognizes and rendersespace
speech/spelling/*
ASR transcripty'a
is now interpreted asi a
speech/date
fixed ASR mis-transcriptions of month namesmai
andjuin
1.17.0 – 2022-04-28
- New and better Voice Activity Detection.
1.16.0 – 2022-04-11
Features:
- More confusions resolved in spelling grammar.
- Punctuation is not enabled by default on legacy grammars anymore (it caused regressions).
- A new spelling grammar with support for punctuation has been added:
builtin:speech/spelling/mixed_with_punct
. - A new regex operator:
||
; it's like|
but it searches till the end of the utterance and it properly enforces precedence of the left-most alternatives.
Fixes:
- Set error status on
RECOGNITION-COMPLETE
when an error occurs.
1.15.0 – 2022-02-07
- New grammar:
builtin:speech/date
, fully functional in French, and partially in European English (numeric dates only in that case).
1.14.0 — 2022-01-26
- Fixed EN support regression
- Support for punctuation and diacritics in spelling grammars
- More robust
START-OF-INPUT
1.13.0 — 2022-01-10
- less sensitive VAD
- text2num fix of "lost" numbers after a silence
- support for custom language extensions
v1.12.4 — 2021-12-08
- better recognition in French spelling for J, G, T, Q, K and 100
v1.12.2 — 2021-11-22
- Fixed Unsigned arithmetic overflow in timers that caused the recognition to hang when input timers are not started and the match is short.
v1.12.1 — 2021-10-25
- Faster no-match conclusion thanks to our
Speech-Nomatch-Timeout
proprietary extension - Improved filtering of noise and low confidence words
- finer heuristics to eliminate false positives at beginning of spelled things
- no more no-match or partial-match returned by hotword (RFC)
- sacrifice a little latency for better accuracy
v1.11.7 — 2021-10-08
- Fixed keyword and boolean grammars to avoid premature rejection of speech
v1.11.5 — 2021-10-04
- Deferred start-of-input
- Fixed boolean serialization
1.11.4 — 2021-10-01
Fixes:
- 'Recognition-Mode' must have priority in all cases
- In the postal address builtin, fixed parts of the city leaking into the street
- Fixed VAD for very low energy signals
- Fixed regression in session handling preventing more than one RECOGNIZE in a session
- Fixed a bug in GET-PARAMS that was returning the wrong value for confidence_threshold
- SET-PARAMS correctly sets default parameters values now, and RECOGNIZE change to those value are not persisted beyond
Improvements:
- better acuracy in most cases for spelling builtins by disabling rescoring
- new buitin
postal_address
deprecates previousaddress
builtin - Internal refactoring and optimizations
- Support legacy hotword parameters in
Vendor-Specific-Parameters
header
v1.6.1 — 2021-06-07
- New Voice Activity Detection algorithm for better accuracy and robustness to background noise (
Sensity-Level
is now ignored). - Improved spelling builtins to distinguish homophones through context (e.g. "de" vs. "deux").
- Improved builtin address with structured interpretation that is now the preferred way to match addresses in production.
v1.5.4 — 2021-04-12
Added a beta version of builtin address with structured interpretation (for French and France only) We'd love feedback if you test it!
Added builtin keywords
Added builtin keywords
v1.5.3 — 2021-03-22
- better session handling
- improved builtin Address (more addresses are recognized)
- fixed filtering of ASR tags
v1.5.0 — 2021-02-08
Features
- Support multiple grammars activation in RECOGNIZE (but not weighted)
- address builtin v1
Fixes
- fixed parasite accented characters in spelling builtins
- fixed letter 'x' transcribed as "ex" in spelling builtins
- fixed premature no-match in normal mode
v1.4.2 — 2021-01-15
- Properly close session after early no match
- Typo fix in nomatch respond body that prevented correct parsing.
v1.4.0 — 2021-01-12
- Adds timing information (
timestamp-start
andtimesramp-end
) in results. - Support
regex
parameter on spelling grammars - Better disambiguation in spelling grammars
v1.3.0 — 2020-12-22
- Normal mode and hotword mode full support according to RFC
- Stricter RFC compliance of timers for Recognition-timeout,
hotword-[min|max]_duration
- Support for statuses
success-maxtime
,hotword-maxtime
,partial-match-maxtime
- Provide interpretation score in ASR score on
<instance>
and<input>
respectively of output payload: - Now return appropriate body for
no-match
andno-input
+ scores interpretation & ASR. - Support for
length
parameter on spelling builtins. - Support for ZipCode builting
v1.2.0 — 2020-09-16
- Bug fix: stream silence while waiting for start of input
- Support for:
- Hotword-Min-Duration (RECOGNIZE method only)
- Hotword-Max-Duration (RECOGNIZE method only)
- Recognition-Mode (RECOGNIZE method only)
- No partial match for spelling grammars
v1.1.0 — 2020-09-10
- Legacy mode
builtin:speech/text2num
to restore behavior of pre-0.13 versions. - Improved
builtin:spelling/letters
to handle double letters said as "2 + letter".
v1.0.0 — 2020-08-24
- English support for plain ASR
- Builtin ASR model configuration
- English support for Spelling builtin
- Log errors with appropriate log priority
- Heartbeat to avoid TCP timeout between RECOGNIZE commands
v0.13.0 — 2020-08-11
- New backend (Voxo-broker), for better scalability and lower latency.
- 60ms bufferization (for lower load)
- New spelling builtin grammar:
- more robust,
- with support for decimal numbers.
- Removed obsolete configuration flag
asr_always_matches
.
Table of content