Rejecto not rejecting?

[mjlee]

Hello – I am using Rejecto with OpenEars, and will be working with speakers of conversational English.. average conversation with many words. Does this work with a vocabulary of around 1000 words? Nearly every time I say a word not in the vocabulary, I get a hypothesis containing words in the vocabulary. In other words, non-vocabulary words are not being rejected at all. For example, if I say just “paraphrase” (not in the vocabulary), I get “CARE” and “AFRAID”.

Thank you in advance for your help!

Here are my settings:
Not using RapidEars
vadThreshold = default
durationOfSilenceBeforeAnalyze = default
rejectoWeight = 2.0;
rejectoUsingVowelsOnly = NO

The vocabulary: (approx 1000 words)

NSArray *firstLanguageArray =
@[@”ACCEPT”,@”FREED”,@”PARTY”,@”ABANDON”,@”ENRAGE”,@”MADDENING”,@”SNOB”,@”ACCEPT”,@”FREEING”,@”PARTY”,@”ABUSE”,@”ENVY”,@”MADDER”,@”SOB”,@”ACCEPTED”,@”FREELY”,@”PASSION”,@”ABUSE”,@”ENVIOUS”,@”MAD”,@”SOBBED”,@”ACCEPTING”,@”PEACE”,@”ACHE”,@”ENVY”,@”MANIAC”,@”SOBBING”,@”ACCEPTS”,@”FREER”,@”PERFECT”,@”ACHING”,@”EVIL”,@”MASOCHIST”,@”SOBS”,@”ACTIVE”,@”FREE”,@”PLAY”,@”ADVERSE”,@”ADVERSARY”,@”EXCRUCIATING”,@”MELANCHOLY”,@”SOLEMN”,@”ADMIRE”,@”ADMIRATION”,@”FRIEND”,@”PLAYED”,@”AFRAID”,@”EXHAUST”,@”MESS”,@”SORROW”,@”ADORE”,@”ADORATION”,@”FUN”,@”PLAYFUL”,@”AGGRAVATE”,@”FAIL”,@”MESSY”,@”SORRY”,@”ADVANTAGE”,@”FUN”,@”PLAYING”,@”AGGRESSIVE”,@”FAKE”,@”MISER”,@”SPITE”,@”ADVENTURE”,@”GENEROUS”,@”PLAYS”,@”AGITATE”,@”FATAL”,@”MISS”,@”STAMMER”,@”AFFECTION”,@”GENTLE”,@”PLEASANT”,@”AGONIZE”,@”FATIGUE”,@”MISSED”,@”STANK”,@”AGREE”,@”GENTLER”,@”PLEASE”,@”AGONY”,@”FAULT”,@”MISSES”,@”STARTLE”,@”STARTLING”,@”AGREEABLE”,@”GENTLEST”,@”PLEASING”,@”ALARM”,@”FEAR”,@”MISSING”,@”STEAL”,@”AGREED”,@”GENTLY”,@”PLEASURE”,@”ALONE”,@”FEARED”,@”MISTAKE”,@”STENCH”,@”AGREEING”,@”GIGGLE”,@”GIGGLING”,@”POPULAR”,@”ANGER”,@”FEARFUL”,@”MOCK”,@”STINK”,@”AGREEMENT”,@”GIVER”,@”POSITIVE”,@”ANGRY”,@”FEARING”,@”MOCKED”,@”STRAIN”,@”AGREES”,@”GIVING”,@”PRAISE”,@”ANGUISH”,@”FEARS”,@”MOCKERY”,@”STRANGE”,@”ALRIGHT”,@”GLAD”,@”PRECIOUS”,@”ANNOY”,@”FEROCIOUS”,@”FEROCITY”,@”MOCKING”,@”STRESS”,@”AMAZE”,@”GLADLY”,@”PRETTY”,@”ANTAGONIZE”,@”ANTAGONISTIC”,@”FEUD”,@”MOCKS”,@”STRUGGLE”,@”AMOROUS”,@”GLAMOR”,@”PRETTY”,@”ANXIETY”,@”ANXIOUS”,@”FIERY”,@”MOLEST”,@”STUBBORN”,@”AMUSE”,@”GLAMOUR”,@”PRIDE”,@”APATHY”,@”APATHETIC”,@”FIGHT”,@”STUNK”,@”GLORY”,@”PRIVILEGE”,@”APPALL”,@”FIRED”,@”MOODY”,@”STUNNED”,@”APPRECIATE”,@”APPRECIATIVE”,@”PRIZE”,@”APPREHENSIVE”,@”FLUNK”,@”MOODY”,@”STUNS”,@”ASSURE”,@”GOOD”,@”PROFIT”,@”FOE”,@”MORON”,@”STUPID”,@”ATTACHMENT”,@”GOODNESS”,@”PROMISE”,@”ARGUE”,@”FOOL”,@”MOURN”,@”STUTTER”,@”ATTRACT”,@”ATTRACTION”,@”GORGEOUS”,@”PROUD”,@”ARROGANT”,@”ARROGANCE”,@”FORBID”,@”MURDER”,@”SUBMISSIVE”,@”AWARD”,@”GRACE”,@”RADIANT”,@”RADIANCE”,@”ASHAME”,@”FOUGHT”,@”NAG”,@”SUCK”,@”AWESOME”,@”GRACED”,@”READINESS”,@”ASSAULT”,@”FRANTIC”,@”NASTY”,@”SUCKED”,@”BEAUTY”,@”BEAUTIFUL”,@”GRACEFUL”,@”READY”,@”ASSHOLE”,@”FREAK”,@”NEEDY”,@”SUCKER”,@”BELOVED”,@”GRACES”,@”REASSURE”,@”ATTACK”,@”FRIGHT”,@”NEGLECT”,@”SUCKS”,@”BENEFICIAL”,@”GRACIOUS”,@”RELAX”,@”ADVERSE”,@”FRUSTRATE”,@”NERD”,@”BENEFIT”,@”GRAND”,@”RELIEF”,@”AVOID”,@”FUCK”,@”NERVOUS”,@”SUFFER”,@”BENEFITS”,@”GRANDE”,@”RELIEVE”,@”AWFUL”,@”FUCKED”,@”NEUROTIC”,@”SUFFERED”,@”BENEFIT”,@”GRATEFUL”,@”RESOLVE”,@”AWKWARD”,@”FUCKER”,@”NUMB”,@”SUFFERER”,@”BENEVOLENT”,@”BENEVOLENCE”,@”GRATITUDE”,@”RESPECT”,@”BAD”,@”FUCKING”,@”OBNOXIOUS”,@”SUFFERING”,@”BENIGN”,@”GREAT”,@”BASHFUL”,@”FUCKS”,@”OBSESS”,@”OBSESSION”,@”SUFFERS”,@”BEST”,@”GRIN”,@”REWARD”,@”BASTARD”,@”FUME”,@”OFFENSE”,@”SUSPICION”,@”SUSPICIOUS”,@”BETTER”,@”GRIN”,@”GRINNING”,@”RICH”,@”BATTLE”,@”FUMING”,@”OFFEND”,@”TANTRUM”,@”BLESS”,@”GRINS”,@”BEATEN”,@”FURIOUS”,@”OFFENSE”,@”TEARS”,@”BOLD”,@”HA”,@”ROMANCE”,@”BITCH”,@”FURY”,@”OUTRAGE”,@”TEASE”,@”BONUS”,@”ROMANTIC”,@”BITTER”,@”GEEK”,@”OVERWHELM”,@”TEMPER”,@”BRAVE”,@”HANDSOME”,@”SAFE”,@”BLAME”,@”GLOOM”,@”PAIN”,@”TEMPERS”,@”BRIGHT”,@”HAPPY”,@”SATISFY”,@”BORE”,@”GODDAM”,@”PAINED”,@”TENSE”,@”BRILLIANT”,@”BRILLIANCE”,@”HAPPY”,@”SAVE”,@”BORING”,@”GOSSIP”,@”PAINFUL”,@”TENSE”,@”CALM”,@”HARMLESS”,@”BOTHER”,@”GRAVE”,@”PAIN”,@”TENSION”,@”CARE”,@”HARMONY”,@”HARMONIOUS”,@”SECURE”,@”BROKE”,@”GREED”,@”PAINS”,@”TERRIBLE”,@”TERRIBLY”,@”CARED”,@”HEARTFELT”,@”SENTIMENTAL”,@”SENTIMENTALITY”,@”BRUTAL”,@”BRUTALITY”,@”GRIEF”,@”PANIC”,@”TERRIFIED”,@”CAREFREE”,@”HEARTWARMING”,@”SHARE”,@”BURDEN”,@”GRIEVE”,@”PARANOID”,@”PARANOIA”,@”TERRIFIES”,@”CAREFUL”,@”HEAVEN”,@”SHARED”,@”CARELESS”,@”GRIM”,@”PATHETIC”,@”TERRIFY”,@”CARES”,@”HEH”,@”SHARES”,@”CHEAT”,@”GROSS”,@”PECULIAR”,@”PECULIARITY”,@”TERRIFYING”,@”CARING”,@”HELPER”,@”SHARING”,@”COMPLAIN”,@”GROUCH”,@”PERVERT”,@”PERVERSE”,@”TERROR”,@”CASUAL”,@”HELPFUL”,@”SILLY”,@”CONFRONT”,@”PESSIMISM”,@”PESSIMIST”,@”THIEF”,@”CASUALLY”,@”HELPING”,@”SILLY”,@”CONFUSE”,@”CONFUSION”,@”GUILT”,@”PETRIFY”,@”CERTAIN”,@”HELPS”,@”SINCERE”,@”SINCERITY”,@”CONTEMPT”,@”CONTEMPTUOUS”,@”HARASS”,@”PETTY”,@”THREAT”,@”CHALLENGE”,@”HERO”,@”SMART”,@”CONTRADICT”,@”HARM”,@”PETTY”,@”TICKED”,@”CHAMP”,@”HILARIOUS”,@”SMILE”,@”CRAP”,@”HARMED”,@”PHOBIA”,@”TIMID”,@”CHARITY”,@”HOHO”,@”SOCIABLE”,@”CRAPPY”,@”HARMFUL”,@”PISS”,@”TORTURE”,@”CHARM”,@”HONEST”,@”CRAZY”,@”HARMING”,@”PITIFUL”,@”TOUGH”,@”CHEER”,@”HONOR”,@”SPECIAL”,@”CRIED”,@”HARMS”,@”PITY”,@”TRAGEDY”,@”CHERISH”,@”HONOR”,@”SPLENDID”,@”CRIES”,@”HATE”,@”POISON”,@”TRAGIC”,@”CHUCKLE”,@”HOPE”,@”STRENGTH”,@”CRITICAL”,@”HATED”,@”PREJUDICE”,@”PREJUDICIAL”,@”TRAUMA”,@”TRAUMATIC”,@”CLEVER”,@”HOPED”,@”STRONG”,@”CRITICIZE”,@”HATEFUL”,@”PRESSURE”,@”TREMBLE”,@”TREMBLING”,@”COMEDY”,@”COMEDIAN”,@”HOPEFUL”,@”SUCCEED”,@”CRUDE”,@”HATER”,@”PRICK”,@”TRICK”,@”COMFORT”,@”HOPEFULLY”,@”SUCCESS”,@”CRUEL”,@”HATES”,@”PROBLEM”,@”TRITE”,@”COMMITMENT”,@”HOPEFULNESS”,@”SUN”,@”CRUSHED”,@”HATING”,@”PROTEST”,@”TRIVIA”,@”TRIVIAL”,@”COMPASSION”,@”HOPES”,@”SUN”,@”CRY”,@”HATRED”,@”PROTESTED”,@”TROUBLE”,@”COMPLIMENT”,@”HOPING”,@”SUN”,@”CRYING”,@”HEARTBREAK”,@”PROTESTING”,@”TURMOIL”,@”CONFIDENCE”,@”HUG”,@”SUNSHINE”,@”CUNT”,@”PUKE”,@”UGH”,@”CONFIDENT”,@”HUG”,@”SUPER”,@”CUT”,@”HEARTLESS”,@”PUNISH”,@”UGLY”,@”CONFIDENTLY”,@”HUGS”,@”SUPERIOR”,@”SUPERIORITY”,@”CYNIC”,@”CYNICISM”,@”HELL”,@”RAGE”,@”UNATTRACTIVE”,@”CONSIDERATE”,@”HUMOR”,@”SUPPORT”,@”DAMAGE”,@”HELLISH”,@”RAGING”,@”UNCERTAIN”,@”CONTENTED”,@”HUMOR”,@”SUPPORTED”,@”DAMN”,@”HELPLESS”,@”RANCID”,@”UNCOMFORTABLE”,@”CONTENTMENT”,@”HURRAH”,@”HURRAY”,@”SUPPORTER”,@”DANGER”,@”HESITATE”,@”HESITATION”,@”RAPE”,@”CONVINCE”,@”IDEAL”,@”SUPPORTING”,@”DAZE”,@”HOMESICK”,@”RAPING”,@”UNEASE”,@”COOL”,@”IMPORTANT”,@”IMPORTANCE”,@”SUPPORTIVE”,@”DECAY”,@”HOPELESS”,@”RAPIST”,@”UNFORTUNATE”,@”COURAGE”,@”COURAGEOUS”,@”IMPRESS”,@”IMPRESSION”,@”SUPPORTS”,@”DEFEAT”,@”HORRIBLE”,@”HORRID”,@”REBEL”,@”REBELLIOUS”,@”UNFRIENDLY”,@”CREATE”,@”IMPROVE”,@”SUPREME”,@”SUPREMACY”,@”DEFECT”,@”DEFECTION”,@”HOSTILE”,@”HOSTILITY”,@”REEK”,@”UNGRATEFUL”,@”CREATIVE”,@”CREATING”,@”IMPROVING”,@”SURE”,@”DEFENCE”,@”HUMILIATE”,@”HUMILIATION”,@”REGRET”,@”UNHAPPY”,@”CREDIT”,@”INCENTIVE”,@”SURPRISE”,@”DEFENSE”,@”HURT”,@”REJECT”,@”REJECTION”,@”UNIMPORTANT”,@”CUTE”,@”INNOCENT”,@”INNOCENCE”,@”SWEET”,@”DEGRADE”,@”IDIOT”,@”RELUCTANT”,@”RELUCTANCE”,@”INSPIRE”,@”SWEETHEART”,@”DEPRESS”,@”DEPRESSION”,@”IGNORE”,@”IGNORANT”,@”REMORSE”,@”UNKIND”,@”DARING”,@”INTELLIGENT”,@”INTELLIGENCE”,@”SWEETIE”,@”DEPRIVED”,@”DEPRIVATION”,@”IMMORAL”,@”IMMORALITY”,@”REPRESS”,@”REPRESSION”,@”DARLIN”,@”DARLING”,@”INTEREST”,@”SWEETLY”,@”DESPAIR”,@”IMPATIENT”,@”IMPATIENCE”,@”RESENT”,@”UNPLEASANT”,@”DEAR”,@”INVIGORATE”,@”SWEETNESS”,@”DESPERATE”,@”DESPERATION”,@”IMPERSONAL”,@”RESIGN”,@”RESIGNATION”,@”UNPROTECTED”,@”DEFINITE”,@”JOKE”,@”SWEETS”,@”DESPISE”,@”IMPOLITE”,@”RESTLESS”,@”UNSAVORY”,@”DEFINITELY”,@”JOKING”,@”TALENT”,@”DESTROY”,@”INADEQUATE”,@”INADEQUACY”,@”REVENGE”,@”UNSUCCESSFUL”,@”DELECTABLE”,@”JOLLY”,@”DESTRUCT”,@”INDECISIVE”,@”RIDICULOUS”,@”RIDICULE”,@”UNSURE”,@”DELICATE”,@”JOY”,@”TENDER”,@”DEVASTATE”,@”INEFFECTUAL”,@”RIGID”,@”UNWELCOME”,@”DELICIOUS”,@”KEEN”,@”TERRIFIC”,@”DEVIL”,@”INFERIOR”,@”INFERIORITY”,@”RISK”,@”UPSET”,@”DELIGHT”,@”KIDDING”,@”THANK”,@”DIFFICULT”,@”INHIBIT”,@”INHIBITION”,@”ROTTEN”,@”UPTIGHT”,@”DETERMINATION”,@”KIND”,@”THANKED”,@”DISADVANTAGE”,@”INSECURE”,@”RUDE”,@”USELESS”,@”DETERMINED”,@”KINDLY”,@”THANKFUL”,@”DISAGREE”,@”INSINCERE”,@”RUIN”,@”VAIN”,@”DEVOTE”,@”DEVOTION”,@”KINDNESS”,@”THANKS”,@”DISAPPOINT”,@”INSULT”,@”SAD”,@”VANITY”,@”DIGNITY”,@”KISS”,@”THOUGHTFUL”,@”DISASTER”,@”INTERRUPT”,@”INTERRUPTION”,@”SADDEN”,@”VICIOUS”,@”DIVINE”,@”THRILL”,@”DISCOMFORT”,@”INTIMIDATE”,@”SADLY”,@”VICTIM”,@”DYNAMIC”,@”DYNAMO”,@”DYNAMISM”,@”LAUGH”,@”TOLERANT”,@”TOLERANCE”,@”DISCOURAGE”,@”IRRATIONAL”,@”SADNESS”,@”VILE”,@”EAGER”,@”LIBERTY”,@”LIBERTINE”,@”TRANQUIL”,@”DISGUST”,@”IRRITATE”,@”IRRITABLE”,@”IRRITABILITY”,@”SARCASM”,@”SARCASTIC”,@”VILLAIN”,@”EASE”,@”LIKE”,@”TREASURE”,@”ISOLATE”,@”ISOLATION”,@”SAVAGE”,@”VIOLATE”,@”VIOLATION”,@”EASIEST”,@”EASIER”,@”LIKEABLE”,@”TREAT”,@”DISILLUSION”,@”JADED”,@”SCARE”,@”VIOLENT”,@”EASILY”,@”LIKED”,@”TRIUMPH”,@”DISLIKE”,@”JEALOUS”,@”SCARING”,@”VULNERABLE”,@”VULNERABILITY”,@”LIKES”,@”TRUE”,@”DISLIKED”,@”JERK”,@”SCARY”,@”VULTURE”,@”EASING”,@”LIKING”,@”DISLIKES”,@”JERKED”,@”SKEPTIC”,@”WAR”,@”EASY”,@”LIVELY”,@”TRUER”,@”DISLIKING”,@”JERKS”,@”SCREAM”,@”WARFARE”,@”ECSTASY”,@”ECSTATIC”,@”TRUEST”,@”DISMAY”,@”KILL”,@”SCREW”,@”WAR”,@”EFFICIENT”,@”EFFICIENCY”,@”TRULY”,@”DISSATISFY”,@”DISSATISFACTION”,@”LAME”,@”SELFISH”,@”WARRING”,@”ELEGANT”,@”ELEGANCE”,@”LOVE”,@”TRUST”,@”DISTRACT”,@”LAZY”,@”SERIOUS”,@”WARS”,@”ENCOURAGE”,@”LOVED”,@”TRUTH”,@”DISTRAUGHT”,@”LAZY”,@”SERIOUSLY”,@”WEAK”,@”ENERGY”,@”ENERGIZE”,@”LOVELY”,@”USEFUL”,@”DISTRESS”,@”LIABILITY”,@”SERIOUSNESS”,@”WEAPON”,@”ENGAGE”,@”LOVER”,@”VALUABLE”,@”DISTRUST”,@”LIAR”,@”SEVERE”,@”WEEP”,@”ENJOY”,@”LOVES”,@”VALUE”,@”DISTURB”,@”LIED”,@”SHAKE”,@”WEIRD”,@”ENTERTAIN”,@”LOVING”,@”VALUED”,@”DOMINATE”,@”DOMINATION”,@”LIES”,@”SHAKY”,@”WEPT”,@”ENTHUSIASM”,@”ENTHUSIASTIC”,@”ENTHUSIAST”,@”LOYAL”,@”VALUES”,@”DOOM”,@”LONE”,@”SHAKY”,@”WHINE”,@”EXCEL”,@”EXCELLENCE”,@”LUCK”,@”VALUING”,@”LONGING”,@”SHAME”,@”WHINING”,@”EXCITE”,@”LUCKED”,@”VIGOR”,@”DOUBT”,@”LOSE”,@”SHIT”,@”WHORE”,@”FAB”,@”LUCKY”,@”VIGOR”,@”DREAD”,@”LOSER”,@”SHOCK”,@”WICKED”,@”FABULOUS”,@”LUCKS”,@”VIRTUE”,@”DULL”,@”LOSES”,@”SHOOK”,@”WIMP”,@”FAITH”,@”LUCKY”,@”VIRTUOUS”,@”VIRTUOSITY”,@”DUMB”,@”LOSING”,@”SHY”,@”WITCH”,@”FANTASTIC”,@”MADLY”,@”VITAL”,@”DUMP”,@”LOSS”,@”SICKEN”,@”WOE”,@”FAVOR”,@”MAGNIFICENT”,@”WARM”,@”DWELL”,@”LOST”,@”SIN”,@”WORRY”,@”FAVOR”,@”MERIT”,@”WEALTH”,@”EGOTIST”,@”LOUSY”,@”LOUSE”,@”SINISTER”,@”WORSE”,@”FEARLESS”,@”MERRY”,@”WELCOME”,@”EMBARRASS”,@”LOW”,@”SINS”,@”WORST”,@”FESTIVE”,@”NEAT”,@”WELL”,@”EMOTIONAL”,@”LUCKLESS”,@”SKEPTIC”,@”WORTHLESS”,@”FIESTA”,@”NICE”,@”WIN”,@”EMPTY”,@”LUDICROUS”,@”SLUT”,@”WRONG”,@”FINE”,@”NURTURE”,@”WIN”,@”ENEMIES”,@”LYING”,@”SMOTHER”,@”YEARN”,@”FLATTER”,@”OK”,@”WINS”,@”ENEMY”,@”MAD”,@”SMUG”,@”FLAWLESS”,@”OKAY”,@”WISDOM”,@”FLEXIBLE”,@”FLEXIBILITY”,@”OKAYS”,@”WISE”,@”FLIRT”,@”WON”,@”FOND”,@”WONDERFUL”,@”FONDLY”,@”OPENNESS”,@”WORSHIP”,@”FONDNESS”,@”OPPORTUNE”,@”WORTHWHILE”,@”FORGAVE”,@”OPTIMAL”,@”WOW”,@”FORGIVE”,@”OPTIMAL”,@”OPTIMIZE”,@”YAY”,@”FREE”,@”ORIGINAL”,@”FREE”,@”OUTGOING”,@”FREEBIE”,@”PAINLESS”,@”PALATABLE”,@”PARADISE”];

Log:

2015-04-29 00:24:45.036 OpenEarsSampleApp[1850:615324] Starting OpenEars logging for OpenEars version 2.03 on 64-bit device (or build): iPhone running iOS version: 8.100000
2015-04-29 00:24:45.037 OpenEarsSampleApp[1850:615324] Creating shared instance of OEPocketsphinxController
2015-04-29 00:24:45.041 OpenEarsSampleApp[1850:615324] Number of words in vocabulary: 970
2015-04-29 00:24:45.371 OpenEarsSampleApp[1850:615324] I’m done running performDictionaryLookup and it took 0.223488 seconds
2015-04-29 00:24:45.385 OpenEarsSampleApp[1850:615324] Starting dynamic language model generation
## Vocab generated by v2 of the CMU-Cambridge Statistcal
## Language Modeling toolkit.
##
## Includes 978 words ##
wfreq2vocab : Done.
text2idngram
Vocab : /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.vocab
Output idngram : /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.idngram
N-gram buffer size : 10
Hash table size : 5000
Temp directory : /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/cmuclmtk-VbUMBA
Max open files : 20
FOF size : 10
n : 3
Initialising hash table…
Reading vocabulary…
Allocating memory for the n-gram buffer…
Reading text into the n-gram buffer…
20,000 n-grams processed for each “.”, 1,000,000 for each line.

Sorting n-grams…
Writing sorted n-grams to temporary file /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/cmuclmtk-VbUMBA/1
Merging 1 temporary files…

2-grams occurring: N times > N times Sug. -spec_num value
0 1953 1982
1 1890 63 73
2 58 5 15
3 4 1 11
4 0 1 11
5 0 1 11
6 0 1 11
7 0 1 11
8 0 1 11
9 0 1 11
10 0 1 11

3-grams occurring: N times > N times Sug. -spec_num value
0 2929 2968
1 2837 92 102
2 86 6 16
3 6 0 10
4 0 0 10
5 0 0 10
6 0 0 10
7 0 0 10
8 0 0 10
9 0 0 10
10 0 0 10
text2idngram : Done.

read_wlist_into_siht: a list of 978 words was read from “/var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.vocab”.
read_wlist_into_array: a list of 978 words was read from “/var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.vocab”.
Unigram was renormalized to absorb a mass of 0.491688
prob[UNK] = 1e-99
ARPA-style 3-gram will be written to /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.arpa
idngram2lm : Done.
INFO: cmd_ln.c(702): Parsing command line:
sphinx_lm_convert \
-i /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.arpa \
-o /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.DMP \
-debug 10

Current configuration:
[NAME] [DEFLT] [VALUE]
-case
-debug 10
-help no no
-i /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.arpa
-ienc
-ifmt
-logbase 1.0001 1.000100e+00
-mmap no no
-o /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.DMP
-oenc utf8 utf8
-ofmt

INFO: ngram_model_arpa.c(504): ngrams 1=978, 2=1952, 3=976
INFO: ngram_model_arpa.c(137): Reading unigrams
INFO: ngram_model_arpa.c(543): 978 = #unigrams created
INFO: ngram_model_arpa.c(197): Reading bigrams
INFO: ngram_model_arpa.c(561): 1952 = #bigrams created
INFO: ngram_model_arpa.c(562): 7 = #prob2 entries
INFO: ngram_model_arpa.c(570): 5 = #bo_wt2 entries
INFO: ngram_model_arpa.c(294): Reading trigrams
INFO: ngram_model_arpa.c(583): 976 = #trigrams created
INFO: ngram_model_arpa.c(584): 4 = #prob3 entries
INFO: ngram_model_dmp.c(518): Building DMP model…
INFO: ngram_model_dmp.c(548): 978 = #unigrams created
INFO: ngram_model_dmp.c(649): 1952 = #bigrams created
INFO: ngram_model_dmp.c(650): 7 = #prob2 entries
INFO: ngram_model_dmp.c(657): 5 = #bo_wt2 entries
INFO: ngram_model_dmp.c(661): 976 = #trigrams created
INFO: ngram_model_dmp.c(662): 4 = #prob3 entries
2015-04-29 00:24:45.502 OpenEarsSampleApp[1850:615324] Done creating language model with CMUCLMTK in 0.116773 seconds.
INFO: cmd_ln.c(702): Parsing command line:
sphinx_lm_convert \
-i /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.arpa \
-o /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.DMP \
-debug 10

Current configuration:
[NAME] [DEFLT] [VALUE]
-case
-debug 10
-help no no
-i /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.arpa
-ienc
-ifmt
-logbase 1.0001 1.000100e+00
-mmap no no
-o /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.DMP
-oenc utf8 utf8
-ofmt

INFO: ngram_model_arpa.c(504): ngrams 1=978, 2=1952, 3=976
INFO: ngram_model_arpa.c(137): Reading unigrams
INFO: ngram_model_arpa.c(543): 978 = #unigrams created
INFO: ngram_model_arpa.c(197): Reading bigrams
INFO: ngram_model_arpa.c(561): 1952 = #bigrams created
INFO: ngram_model_arpa.c(562): 9 = #prob2 entries
INFO: ngram_model_arpa.c(570): 5 = #bo_wt2 entries
INFO: ngram_model_arpa.c(294): Reading trigrams
INFO: ngram_model_arpa.c(583): 976 = #trigrams created
INFO: ngram_model_arpa.c(584): 5 = #prob3 entries
INFO: ngram_model_dmp.c(518): Building DMP model…
INFO: ngram_model_dmp.c(548): 978 = #unigrams created
INFO: ngram_model_dmp.c(649): 1952 = #bigrams created
INFO: ngram_model_dmp.c(650): 9 = #prob2 entries
INFO: ngram_model_dmp.c(657): 5 = #bo_wt2 entries
INFO: ngram_model_dmp.c(661): 976 = #trigrams created
INFO: ngram_model_dmp.c(662): 5 = #prob3 entries
2015-04-29 00:24:45.688 OpenEarsSampleApp[1850:615324] I’m done running dynamic language model generation and it took 0.646351 seconds
2015-04-29 00:24:45.689 OpenEarsSampleApp[1850:615324] Attempting to start listening session from startListeningWithLanguageModelAtPath:
2015-04-29 00:24:45.694 OpenEarsSampleApp[1850:615324] User gave mic permission for this app.
2015-04-29 00:24:45.694 OpenEarsSampleApp[1850:615324] Valid setSecondsOfSilence value of 0.700000 will be used.
2015-04-29 00:24:45.695 OpenEarsSampleApp[1850:615324] Successfully started listening session from startListeningWithLanguageModelAtPath:
2015-04-29 00:24:45.695 OpenEarsSampleApp[1850:615345] Starting listening.
2015-04-29 00:24:45.695 OpenEarsSampleApp[1850:615345] about to set up audio session
2015-04-29 00:24:45.819 OpenEarsSampleApp[1850:615349] Audio route has changed for the following reason:
2015-04-29 00:24:46.046 OpenEarsSampleApp[1850:615345] done starting audio unit
INFO: cmd_ln.c(702): Parsing command line:
\
-lm /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.DMP \
-vad_prespeech 10 \
-vad_postspeech 69 \
-vad_threshold 2.000000 \
-remove_noise yes \
-remove_silence yes \
-bestpath no \
-lw 6.500000 \
-dict /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.dic \
-hmm /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle

Current configuration:
[NAME] [DEFLT] [VALUE]
-agc none none
-agcthresh 2.0 2.000000e+00
-allphone
-allphone_ci no no
-alpha 0.97 9.700000e-01
-argfile
-ascale 20.0 2.000000e+01
-aw 1 1
-backtrace no no
-beam 1e-48 1.000000e-48
-bestpath yes no
-bestpathlw 9.5 9.500000e+00
-bghist no no
-ceplen 13 13
-cmn current current
-cmninit 8.0 8.0
-compallsen no no
-debug 0
-dict /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.dic
-dictcase no no
-dither no no
-doublebw no no
-ds 1 1
-fdict
-feat 1s_c_d_dd 1s_c_d_dd
-featparams
-fillprob 1e-8 1.000000e-08
-frate 100 100
-fsg
-fsgusealtpron yes yes
-fsgusefiller yes yes
-fwdflat yes yes
-fwdflatbeam 1e-64 1.000000e-64
-fwdflatefwid 4 4
-fwdflatlw 8.5 8.500000e+00
-fwdflatsfwin 25 25
-fwdflatwbeam 7e-29 7.000000e-29
-fwdtree yes yes
-hmm /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle
-input_endian little little
-jsgf
-kdmaxbbi -1 -1
-kdmaxdepth 0 0
-kdtree
-keyphrase
-kws
-kws_plp 1e-1 1.000000e-01
-kws_threshold 1 1.000000e+00
-latsize 5000 5000
-lda
-ldadim 0 0
-lextreedump 0 0
-lifter 0 0
-lm /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.DMP
-lmctl
-lmname
-logbase 1.0001 1.000100e+00
-logfn
-logspec no no
-lowerf 133.33334 1.333333e+02
-lpbeam 1e-40 1.000000e-40
-lponlybeam 7e-29 7.000000e-29
-lw 6.5 6.500000e+00
-maxhmmpf 10000 10000
-maxnewoov 20 20
-maxwpf -1 -1
-mdef
-mean
-mfclogdir
-min_endfr 0 0
-mixw
-mixwfloor 0.0000001 1.000000e-07
-mllr
-mmap yes yes
-ncep 13 13
-nfft 512 512
-nfilt 40 40
-nwpen 1.0 1.000000e+00
-pbeam 1e-48 1.000000e-48
-pip 1.0 1.000000e+00
-pl_beam 1e-10 1.000000e-10
-pl_pbeam 1e-5 1.000000e-05
-pl_window 0 0
-rawlogdir
-remove_dc no no
-remove_noise yes yes
-remove_silence yes yes
-round_filters yes yes
-samprate 16000 1.600000e+04
-seed -1 -1
-sendump
-senlogdir
-senmgau
-silprob 0.005 5.000000e-03
-smoothspec no no
-svspec
-tmat
-tmatfloor 0.0001 1.000000e-04
-topn 4 4
-topn_beam 0 0
-toprule
-transform legacy legacy
-unit_area yes yes
-upperf 6855.4976 6.855498e+03
-usewdphones no no
-uw 1.0 1.000000e+00
-vad_postspeech 50 69
-vad_prespeech 10 10
-vad_threshold 2.0 2.000000e+00
-var
-varfloor 0.0001 1.000000e-04
-varnorm no no
-verbose no no
-warp_params
-warp_type inverse_linear inverse_linear
-wbeam 7e-29 7.000000e-29
-wip 0.65 6.500000e-01
-wlen 0.025625 2.562500e-02

2015-04-29 00:24:46.052 OpenEarsSampleApp[1850:615349] There was a category change. The new category is AVAudioSessionCategoryPlayAndRecord
INFO: cmd_ln.c(702): Parsing command line:
\
-nfilt 25 \
-lowerf 130 \
-upperf 6800 \
-feat 1s_c_d_dd \
-svspec 0-12/13-25/26-38 \
-agc none \
-cmn current \
-varnorm no \
-transform dct \
-lifter 22 \
-cmninit 40

Current configuration:
[NAME] [DEFLT] [VALUE]
-agc none none
-agcthresh 2.0 2.000000e+00
-alpha 0.97 9.700000e-01
-ceplen 13 13
-cmn current current
-cmninit 8.0 40
-dither no no
-doublebw no no
-feat 1s_c_d_dd 1s_c_d_dd
-frate 100 100
-input_endian little little
-lda
-ldadim 0 0
-lifter 0 22
-logspec no no
-lowerf 133.33334 1.300000e+02
-ncep 13 13
-nfft 512 512
-nfilt 40 25
-remove_dc no no
-remove_noise yes yes
-remove_silence yes yes
-round_filters yes yes
-samprate 16000 1.600000e+04
-seed -1 -1
-smoothspec no no
-svspec 0-12/13-25/26-38
-transform legacy dct
-unit_area yes yes
-upperf 6855.4976 6.800000e+03
-vad_postspeech 50 69
-vad_prespeech 10 10
-vad_threshold 2.0 2.000000e+00
-varnorm no no
-verbose no no
-warp_params
-warp_type inverse_linear inverse_linear
-wlen 0.025625 2.562500e-02

INFO: acmod.c(252): Parsed model-specific feature parameters from /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/feat.params
INFO: feat.c(715): Initializing feature stream to type: ‘1s_c_d_dd’, ceplen=13, CMN=’current’, VARNORM=’no’, AGC=’none’
INFO: cmn.c(143): mean[0]= 12.00, mean[1..12]= 0.0
INFO: acmod.c(171): Using subvector specification 0-12/13-25/26-38
INFO: mdef.c(518): Reading model definition: /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/mdef
2015-04-29 00:24:46.070 OpenEarsSampleApp[1850:615349] This is not a case in which OpenEars notifies of a route change. At the close of this function, the new audio route is —SpeakerMicrophoneBuiltIn—. The previous route before changing to this route was <AVINFO: mdef.c(531): Found byte-order mark BMDF, assuming this is a binary mdef file
INFO: bin_mdef.c(336): Reading binary model definition: /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/mdef
AudioSessionRouteDescription: 0x174603720,
inputs = (null);
outputs = (
“<AVAudioSessionPortDescription: 0x174603750, type = Speaker; name = Speaker; UID = Speaker; selectedDataSource = (null)>”
)>.
INFO: bin_mdef.c(516): 46 CI-phone, 168344 CD-phone, 3 emitstate/phone, 138 CI-sen, 6138 Sen, 32881 Sen-Seq
INFO: tmat.c(206): Reading HMM transition probability matrices: /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/transition_matrices
INFO: acmod.c(124): Attempting to use SCHMM computation module
INFO: ms_gauden.c(198): Reading mixture gaussian parameter: /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/means
INFO: ms_gauden.c(292): 1 codebook, 3 feature, size:
INFO: ms_gauden.c(294): 512×13
INFO: ms_gauden.c(294): 512×13
INFO: ms_gauden.c(294): 512×13
INFO: ms_gauden.c(198): Reading mixture gaussian parameter: /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/variances
INFO: ms_gauden.c(292): 1 codebook, 3 feature, size:
INFO: ms_gauden.c(294): 512×13
INFO: ms_gauden.c(294): 512×13
INFO: ms_gauden.c(294): 512×13
INFO: ms_gauden.c(354): 0 variance values floored
INFO: s2_semi_mgau.c(904): Loading senones from dump file /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/sendump
INFO: s2_semi_mgau.c(928): BEGIN FILE FORMAT DESCRIPTION
INFO: s2_semi_mgau.c(991): Rows: 512, Columns: 6138
INFO: s2_semi_mgau.c(1023): Using memory-mapped I/O for senones
INFO: s2_semi_mgau.c(1294): Maximum top-N: 4 Top-N beams: 0 0 0
INFO: dict.c(320): Allocating 5197 * 32 bytes (162 KiB) for word entries
INFO: dict.c(333): Reading main dictionary: /var/mobile/Containers/Data/Application/81636750-54D2-4106-A28D-7332A254D365/Library/Caches/LangModel.dic
INFO: dict.c(213): Allocated 7 KiB for strings, 11 KiB for phones
INFO: dict.c(336): 1092 words read
INFO: dict.c(342): Reading filler dictionary: /private/var/mobile/Containers/Bundle/Application/789ABCF3-0765-4A02-AB6F-D875F4F7033C/OpenEarsSampleApp.app/AcousticModelEnglish.bundle/noisedict
INFO: dict.c(213): Allocated 0 KiB for strings, 0 KiB for phones
INFO: dict.c(345): 9 words read
INFO: dict2pid.c(396): Building PID tables for dictionary
INFO: dict2pid.c(406): Allocating 46^3 * 2 bytes (190 KiB) for word-initial triphones
INFO: dict2pid.c(132): Allocated 51152 bytes (49 KiB) for word-final triphones
INFO: dict2pid.c(196): Allocated 51152 bytes (49 KiB) for single-phone word triphones
INFO: ngram_model_arpa.c(79): No \data\ mark in LM file
INFO: ngram_model_dmp.c(166): Will use memory-mapped I/O for LM file
INFO: ngram_model_dmp.c(220): ngrams 1=978, 2=1952, 3=976
INFO: ngram_model_dmp.c(266): 978 = LM.unigrams(+trailer) read
INFO: ngram_model_dmp.c(312): 1952 = LM.bigrams(+trailer) read
INFO: ngram_model_dmp.c(338): 976 = LM.trigrams read
INFO: ngram_model_dmp.c(363): 9 = LM.prob2 entries read
INFO: ngram_model_dmp.c(383): 5 = LM.bo_wt2 entries read
INFO: ngram_model_dmp.c(403): 5 = LM.prob3 entries read
INFO: ngram_model_dmp.c(431): 4 = LM.tseg_base entries read
INFO: ngram_model_dmp.c(487): 978 = ascii word strings read
INFO: ngram_search_fwdtree.c(99): 245 unique initial diphones
INFO: ngram_search_fwdtree.c(148): 0 root, 0 non-root channels, 49 single-phone words
INFO: ngram_search_fwdtree.c(186): Creating search tree
INFO: ngram_search_fwdtree.c(192): before: 0 root, 0 non-root channels, 49 single-phone words
INFO: ngram_search_fwdtree.c(326): after: max nonroot chan increased to 3047
INFO: ngram_search_fwdtree.c(339): after: 245 root, 2919 non-root channels, 48 single-phone words
INFO: ngram_search_fwdflat.c(157): fwdflat: min_ef_width = 4, max_sf_win = 25
2015-04-29 00:24:46.165 OpenEarsSampleApp[1850:615345] There was no previous CMN value in the plist so we are using the fresh CMN value 42.000000.
2015-04-29 00:24:46.166 OpenEarsSampleApp[1850:615345] Listening.
2015-04-29 00:24:46.168 OpenEarsSampleApp[1850:615345] Project has these words or phrases in its dictionary:
___REJ_AA
___REJ_AE
___REJ_AH
___REJ_AO
___REJ_AW
___REJ_AY
___REJ_B
___REJ_CH
___REJ_D
___REJ_DH
___REJ_EH
___REJ_ER
___REJ_EY
___REJ_F
___REJ_G
___REJ_HH
___REJ_IH
___REJ_IY
___REJ_JH
___REJ_K
___REJ_L
___REJ_M
___REJ_N
___REJ_NG
___REJ_OW
___REJ_OY
___REJ_P
___REJ_R
___REJ_S
___REJ_SH
___REJ_T
…and 1062 more.
2015-04-29 00:24:46.169 OpenEarsSampleApp[1850:615345] Recognition loop has started
2015-04-29 00:24:46.233 OpenEarsSampleApp[1850:615324] Local callback: Pocketsphinx is now listening.
2015-04-29 00:24:46.235 OpenEarsSampleApp[1850:615324] Local callback: Pocketsphinx started.
2015-04-29 00:24:47.815 OpenEarsSampleApp[1850:615345] Speech detected…
INFO: ngram_search.c(462): Resized backpointer table to 10000 entries
2015-04-29 00:24:49.605 OpenEarsSampleApp[1850:615345] End of speech detected…
INFO: cmn_prior.c(131): cmn_prior_update: from < 42.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 >
INFO: cmn_prior.c(149): cmn_prior_update: to < 38.09 0.20 -10.59 2.18 2.53 -5.46 -8.91 -10.75 3.32 -0.04 -5.32 6.43 2.00 >
INFO: ngram_search_fwdtree.c(1550): 6261 words recognized (33/fr)
INFO: ngram_search_fwdtree.c(1552): 425506 senones evaluated (2240/fr)
INFO: ngram_search_fwdtree.c(1556): 288250 channels searched (1517/fr), 45570 1st, 52353 last
INFO: ngram_search_fwdtree.c(1559): 10175 words for which last channels evaluated (53/fr)
INFO: ngram_search_fwdtree.c(1561): 11847 candidate words for entering last phone (62/fr)
INFO: ngram_search_fwdtree.c(1564): fwdtree 0.40 CPU 0.210 xRT
INFO: ngram_search_fwdtree.c(1567): fwdtree 3.21 wall 1.690 xRT
INFO: ngram_search_fwdflat.c(302): Utterance vocabulary contains 96 words
INFO: ngram_search_fwdflat.c(938): 3622 words recognized (19/fr)
INFO: ngram_search_fwdflat.c(940): 101319 senones evaluated (533/fr)
INFO: ngram_search_fwdflat.c(942): 96622 channels searched (508/fr)
INFO: ngram_search_fwdflat.c(944): 10499 words searched (55/fr)
INFO: ngram_search_fwdflat.c(947): 5903 word transitions (31/fr)
INFO: ngram_search_fwdflat.c(950): fwdflat 0.15 CPU 0.081 xRT
INFO: ngram_search_fwdflat.c(953): fwdflat 0.15 wall 0.080 xRT
2015-04-29 00:24:49.761 OpenEarsSampleApp[1850:615345] Pocketsphinx heard “CARE AFRAID” with a score of (0) and an utterance ID of 0.
2015-04-29 00:24:49.762 OpenEarsSampleApp[1850:615324] Local callback: The received hypothesis is CARE AFRAID with a score of 0 and an ID of 0

[mjlee]

Additional info:
Spoken words that are in the vocabulary are correctly being identified.
There is almost no background noise, and I am speaking with the device near my mouth, as if I’m talking on the phone.
Nearly every non-vocabulary word I say is not rejected, and recognized as a vocabulary word.

[Halle Winkler]

Welcome,

Yes, sorry, I would say that this is expected behavior for a vocabulary so large that there are several perfect rhyming matches for OOV (out of vocabulary) utterances in the vocabulary, when the utterances are being spoken directly into the device at the same level as the rest of the interaction. If it were possible for Rejecto to be so highly weighted that it would override a large vocabulary’s ability to match these utterances, there would be too much confusion to confidently identify in-vocabulary utterances as well. This is actually the same with something like Siri – if you insert an OOV word directly into an otherwise-contextually-understandable Siri interaction (e.g. a name Siri doesn’t know, a foreign word, or something like medical jargon which isn’t in the relevant vocab) it will either be transcribed into something else or the question will be thrown out.

It has only been in the last couple of device versions that a 1000-word vocabulary could be reasonably made use of in OpenEars, so Rejecto is more geared towards use with a smaller vocabulary which has some “holes” in it in terms of not having multiple rhyming matches for the majority of OOV utterances.

Is this a realistic interaction for your app, that there is clear speech directed at it intentionally under ideal circumstances but the speech is unknown to the app? Generally OOV rejection is more optimized for tuning out speech which isn’t intended for the app, ruling out non-speech sounds rather than submitting them for hypotheses, and otherwise avoiding input of a more accidental nature. Users directing a lot of clear intentional speech towards the app that the app doesn’t know about might more come under the category of user education requirements than something needing a technological fix – it depends a bit on how you see this occurring.

Regarding the very large vocabulary, do you know that you can switch between smaller vocabularies almost instantly? So if your requirement is to present a sentence the user is supposed to say correctly, you can just attempt to recognize those words plus the Rejecto phonemes rather than the entire vocabulary at once.

[mjlee]

Thanks for the fast reply. I tried cutting it down to about 200 words, with similar thought somewhat better results. From what I read about Rejecto, it seemed that it is designed to reject words that are OOV. But it sounds like what it’s really designed for is rejecting non-word utterances? Switching to a small vocabulary won’t work in my case, as my goal is to let the user say anything, and detect only certain words and reject all others. It sounds like OpenEars with or without Rejecto is probably not a good solution for what I’m trying to do?

[Halle Winkler]

Rejecto is fine at rejecting OOV utterances, but it isn’t going to work ideally in a language model is large enough that it has several other perfect rhymes for the OOV utterances if they are being spoken directly to the app, since they will compete quite evenly with Rejecto’s rejection. I would be surprised if the results were the same with 200 words, perhaps you should create a replication case to show me if you have time:

https://www.politepix.com/forums/topic/how-to-create-a-minimal-case-for-replication/

It is possible that an app concept that is based on the user saying any possible sentence and then a very large vocabulary of keywords being detected within that isn’t a good fit for offline recognition. This is basically a variation on free-form dictation, so if you can’t make the vocabulary small enough that it isn’t full of rhyming combinations for OOV words and you can’t restrict the search space of the input, it will probably work better to use a cloud-based large vocabulary dictation SDK instead of an offline speech UI SDK, and just identify and then process everything.

[Author]

Posts