Examples of edu.cmu.sphinx.result.SausageMaker

edu.cmu.sphinx.result.SausageMaker

The SausageMaker takes word lattices as input and turns them into sausages (Confusion Networks) according to Mangu, Brill and Stolcke, "Finding Consensus in Speech Recognition: word error minimization and other applications of confusion networks", Computer Speech and Language, 2000. Note that the getBestHypothesis of the ConfidenceResult object returned by the {@link #score(Result) score} method returns the path where all the words havethe highest posterior probability within its corresponding time slot.

@author pgorniak

        if (result.isFinal() && ref != null) {
            Lattice lattice = new Lattice(result);
            LatticeOptimizer optimizer = new LatticeOptimizer(lattice);
            optimizer.optimize();           
            lattice.computeNodePosteriors(languageModelWeight);
            SausageMaker sausageMaker = new SausageMaker(lattice);
            Sausage sausage = sausageMaker.makeSausage();
            sausage.removeFillers();
                
            getAligner().alignSausage(ref, sausage);
            showFullPath(result);
            showDetails(result.toString());

Examples of edu.cmu.sphinx.result.SausageMaker

Related Classes of edu.cmu.sphinx.result.SausageMaker