Method of setting optimum-partitioned classified neural network and method and apparatus for automatic labeling using optimum-partitioned classified neural network
First Claim
1. A method of setting an optimum-partitioned classified neural network, comprising:
- obtaining a number of L phoneme combinations having names of left and right phonemes by using a phoneme boundary obtained by manual labeling;
generating a number of K neural network combinations of an MLP type from learning data including input variables;
searching for neural networks having minimum errors with respect to the L phoneme combinations from the neural network combinations, and classifying the L phoneme combinations into K phoneme combination groups searched with the same neural networks;
using the K phoneme combination groups classified in the searching for neural networks, updating weights until individual errors of the neural networks have converged during learning with applicable learning data for the K neural networks; and
repeatedly performing the searching for neural networks and the updating weight, corresponding to the K neural networks of which the individual errors have converged, until a total error sum of K neural networks, of which individual errors have converged in the updating weights, has converged, and composing an optimum-partitioned classified neural network combination using the K neural networks obtained when the total error sum has converged.
1 Assignment
0 Petitions
Accused Products
Abstract
A method of automatic labeling using an optimum-partitioned classified neural network includes searching for neural networks having minimum errors with respect to a number of L phoneme combinations from a number of K neural network combinations generated at an initial stage or updated, updating weights during learning of the K neural networks by K phoneme combination groups searched with the same neural networks, and composing an optimum-partitioned classified neural network combination using the K neural networks of which a total error sum has converged; and tuning a phoneme boundary of a first label file by using the phoneme combination group classification result and the optimum-partitioned classified neural network combination, and generating a final label file reflecting the tuning result.
46 Citations
41 Claims
-
1. A method of setting an optimum-partitioned classified neural network, comprising:
-
obtaining a number of L phoneme combinations having names of left and right phonemes by using a phoneme boundary obtained by manual labeling;
generating a number of K neural network combinations of an MLP type from learning data including input variables;
searching for neural networks having minimum errors with respect to the L phoneme combinations from the neural network combinations, and classifying the L phoneme combinations into K phoneme combination groups searched with the same neural networks;
using the K phoneme combination groups classified in the searching for neural networks, updating weights until individual errors of the neural networks have converged during learning with applicable learning data for the K neural networks; and
repeatedly performing the searching for neural networks and the updating weight, corresponding to the K neural networks of which the individual errors have converged, until a total error sum of K neural networks, of which individual errors have converged in the updating weights, has converged, and composing an optimum-partitioned classified neural network combination using the K neural networks obtained when the total error sum has converged. - View Dependent Claims (2, 12)
-
-
3. A method of automatic labeling to tune a phoneme boundary of a first label file generated by performing automatic labeling of a manual label file, the method comprising:
-
searching for neural networks having minimum errors with respect to a number of L phoneme combinations from a number of K neural network combinations generated at an initial stage or updated;
updating weights during learning of the K neural networks by K phoneme combination groups searched with the same neural networks;
composing an optimum-partitioned classified neural network combination using the K neural networks of which a total error sum has converged;
tuning a phoneme boundary of a first label file using a phoneme combination group classification result and the optimum-partitioned classified neural network combination from the composing the optimum-partitioned classified neural network combination; and
generating a final label file reflecting the tuning result. - View Dependent Claims (4, 5, 6, 13)
-
-
7. An apparatus for automatic labeling using an optimum-partitioned classified neural network, comprising:
-
a labeling unit to generate a first label file by performing automatic labeling for a manual label file;
an optimum-partitioned classified neural network composing unit searching neural networks having minimum errors with respect to a number of L phoneme combinations from a number of K neural network combinations generated at an initial stage or updated, updating weights during learning of the K neural networks by K phoneme combination groups searched with the same neural networks, and composing an optimum-partitioned classified neural network combination using the K neural networks of which a total error sum has converged; and
a phoneme boundary tuning unit tuning a phoneme boundary of the first label file by using a phoneme combination group classification result and the optimum-partitioned classified neural network combination supplied from the optimum-partitioned classified neural network composing unit, and generating a final label file reflecting the tuning result. - View Dependent Claims (8, 9, 10, 11)
-
-
14. An apparatus for automatic labeling using an optimum-partitioned classified neural network, comprising:
-
a labeling unit to perform automatic labeling of a manual label file and generate a first label file;
an optimum-partitioned classified neural network composing unit to receive input variables, segment phoneme combinations into partitions applicable to neural networks, and compose optimum-partitioned classified neural networks of Multi-Layer Perceptron-type from re-learned partitions; and
a phoneme boundary tuning unit to tune a phoneme boundary of the first label file supplied from the labeling unit and to generate a final label file reflecting the tuning result. - View Dependent Claims (15, 16, 17)
-
-
18. A method of composing the optimum neural network combination minimizing a total error sum, comprising:
-
preparing an initial neural network combination using input variables;
searching for the optimum neural network having minimum errors with respect to the phoneme combinations from the initial neural network combination and classifying phoneme combinations with optimum neural networks;
merging phoneme combinations with the same neural network and classifying the merged phoneme combinations into new partitions;
updating each neural network and learning each neural network according to the partitions generated by the classifying and the merging;
determining whether all neural networks are converged, wherein if all neural networks are converged, composing the neural network combination as an optimum-partitioned classified neural network combination, and if all neural networks are not converged, then re-learning the optimum neural network having minimum by repeating the classifying, the merging, and the updating operations until all neural networks are converged. - View Dependent Claims (19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37)
-
-
38. A method of learning and updating a neural network, comprising:
-
preparing initial neural network combinations composing neural networks of multi-layer perceptron-type;
searching a multi-layer perceptron index having a minimum error in the initial neural network combinations of all phoneme combinations;
classifying the merged phoneme combinations into new partitions by merging phoneme combinations with the same multi-layer perceptron index if the multi-layer perceptron index having a minimum error for all phoneme combinations is searched; and
re-training neural networks to update weights by learning data applicable to each partition, wherein the re-training procedure of individual neural networks calculates errors using the updated weights and repeating the re-training until a changing rate of errors becomes smaller than a first threshold value. - View Dependent Claims (39, 40)
-
-
41. A method of operating an optimum-partitioned classified neural network, comprising:
-
obtaining phoneme combinations composed with names of left and right phonemes by using a phoneme boundary obtained by manual labeling;
generating neural network combinations of an Multi-Layer Perceptron-type from learning data including input variables;
searching for neural networks having minimum errors with respect the phoneme combinations from the neural network combinations, and classifying the phoneme combinations into phoneme combination groups searched with the same neural networks;
using the phoneme combination groups classified in the searching for neural networks, updating weights until individual errors of the neural networks have converged during learning with applicable learning data for the neural networks; and
repeatedly performing the searching for neural networks and the updating weight, corresponding to the neural networks of which the individual errors have converged, until a total error sum of the neural networks, of which individual errors have converged in the updating weights, has converged, and composing an optimum-partitioned classified neural network combination using the K neural networks obtained when the total error sum has converged.
-
Specification