Package edu.cmu.sphinx.util

Examples of edu.cmu.sphinx.util.TimeFrame

103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
        Queue<List<String>> texts = new ArrayDeque<List<String>>();
        Queue<TimeFrame> timeFrames = new ArrayDeque<TimeFrame>();

        ranges.offer(new Range(0, transcript.size()));
        texts.offer(transcript);
        TimeFrame totalTimeFrame = TimeFrame.INFINITE;
        timeFrames.offer(totalTimeFrame);
        long lastFrame = TimeFrame.INFINITE.getEnd();

        for (int i = 0; i < 4; ++i) {
            if (i == 3) {
                context.setLocalProperty("decoder->searchManager",
                        "alignerSearchManager");
                context.processBatch();
            }

            while (!texts.isEmpty()) {
                assert texts.size() == ranges.size();
                assert texts.size() == timeFrames.size();

                List<String> text = texts.poll();
                TimeFrame frame = timeFrames.poll();
                Range range = ranges.poll();

                logger.info("Aligning frame " + frame + " to text " + text
                        + " range " + range);
View Full Code Here
245
246
247
248
249
250
251
252
                    + transcript.subList(start, end).toString());
            return;
        }

        texts.offer(transcript.subList(start, end));
        timeFrames.offer(new TimeFrame(timeStart, timeEnd));
        ranges.offer(new Range(start, end - 1));
    }
View Full Code Here
74
75
76
77
78
79
80
81
     * @param node the node to extract information from
     * @param confidence the confidence (posterior) to assign
     */
    public WordResult(Node node, double confidence) {
        this(node.getWord(),
             new TimeFrame(node.getBeginTime(), node.getEndTime()),
             LogMath.LOG_ZERO, confidence);
    }
View Full Code Here
79
80
81
82
83
84
85
86
87
88
89
90
91
        for (List<WordResult> utt : utts) {
            long startFrame = Long.MAX_VALUE;
            long endFrame = Long.MIN_VALUE;

            for (WordResult result : utt) {
                TimeFrame frame = result.getTimeFrame();
                startFrame = Math.min(startFrame, frame.getStart());
                endFrame = Math.max(endFrame, frame.getEnd());
                System.out.print(result.getPronunciation().getWord());
                System.out.print(' ');
            }

            String[] basename = inFile.getName().split("\\.wav$");
View Full Code Here
409
410
411
412
413
414
415
416
417
418
419
420
            while (token != null) {
                if (token.isWord()) {
                    Word word = token.getWord();
                    if (withFillers || !word.isFiller()) {
                        TimeFrame timeFrame =
                                new TimeFrame(
                                        ((FloatData) prevFeature)
                                                .getCollectTime(),
                                        ((FloatData) prevWordFirstFeature)
                                                .getCollectTime());
                        result.add(new WordResult(word, timeFrame, token
View Full Code Here
449
450
451
452
453
454
455
456
457
458
459
        while (token != null) {
            if (token.isWord()) {
                if (word != null && lastFeature != null) {
                    if (withFillers || !word.isFiller()) {
                        TimeFrame timeFrame = new TimeFrame(((FloatData) lastFeature).getCollectTime(),
                                ((FloatData) lastWordFirstFeature).getCollectTime());
                        result.add(new WordResult(word, timeFrame, token.getScore(), 1.));
                    }
                    word = token.getWord();
                    lastWordFirstFeature = lastFeature;
View Full Code Here
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
                .setLanguageModelPath("resource:/edu/cmu/sphinx/models/language/en-us.lm.dmp");

        StreamSpeechRecognizer recognizer = new StreamSpeechRecognizer(
                configuration);

        TimeFrame t;
        SpeechResult result;

        for (SpeakerCluster spk : speakers) {
            Stats stats = recognizer.createStats(1);
            ArrayList<Segment> segments = spk.getSpeakerIntervals();

            for (Segment s : segments) {
                long startTime = s.getStartTime();
                long endTime = s.getStartTime() + s.getLength();
                t = new TimeFrame(startTime, endTime);

                recognizer.startRecognition(url.openStream(), t);
                while ((result = recognizer.getResult()) != null) {
                    stats.collect(result);
                }
                recognizer.stopRecognition();
            }

            Transform profile;
            // Create the Transformation
            profile = stats.createTransform();
            recognizer.setTransform(profile);

            for (Segment seg : segments) {
                long startTime = seg.getStartTime();
                long endTime = seg.getStartTime() + seg.getLength();
                t = new TimeFrame(startTime, endTime);

                // Decode again with updated SpeakerProfile
                recognizer.startRecognition(url.openStream(), t);
                while ((result = recognizer.getResult()) != null) {
                    System.out.format("Hypothesis: %s\n",
View Full Code Here
TOP

Related Classes of edu.cmu.sphinx.util.TimeFrame

  • edu.cmu.sphinx.api.SpeechAligner
  • edu.cmu.sphinx.demo.speakerid.SpeakerIdentificationDemo
  • edu.cmu.sphinx.result.Result
  • edu.cmu.sphinx.result.WordResult
  • edu.cmu.sphinx.tools.aligner.Aligner
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.