Pattern recognition device, pattern recognition method and computer program product

A computer program product having a computer readable medium including programmed instructions that, when executed by a computer, cause the computer to perform:

• calculating a feature vector from input pattern;
• selecting, from models, one or more selection models whose distances from the feature vector are within a predetermined range, each of the models being represented by both an average vector of a pattern and a variance vector indicating variance of the pattern, and the average vector and the variance vector each having N dimensional components (N≧1);
• calculating, for each of the selection models, a modified directional vector having the N dimensional components, a value of an n-th dimensional component (N≧n≧1) of the modified directional vector being obtained by subtracting a value of the n-th dimensional component of the variance vector multiplied by a predetermined coefficient from an absolute value of the n-th component of a different vector between the average vector and the feature vector to obtain a first value, and then multiplying the first value by a plus or minus sign identical to a sign of the n-th dimensional component of the difference vector, and further calculate a correction vector with respect to a vector obtained by superimposing the modified directional vectors;
• correcting the feature vector by using the correction vector;
• and
• performing recognition of the input pattern by using the feature vector corrected.

A pattern recognition method comprising:

• calculating, by a feature vector calculator, a feature vector from input pattern;
• selecting by a model selecting unit, from models, one or more selection models whose distances from the feature vector are within a predetermined range, each of the models being represented by both an average vector of a pattern and a variance vector indicating variance of the pattern, and the average vector and the variance vector each having N dimensional components (N≧1);
• calculating by a correction vector calculator, for each of the selection models, a modified directional vector having the N dimensional components, a value of an n-th dimensional component (N≧n≧1) of the modified directional vector being obtained by subtracting a value of the n-th dimensional component of the variance vector multiplied by a predetermined coefficient from an absolute value of the n-th component of a different vector between the average vector and the feature vector to obtain a first value, and then multiplying the first value by a plus or minus sign identical to a sign of the n-th dimensional component of the difference vector, and further calculate a correction vector with respect to a vector obtained by superimposing the modified directional vectors;
• correcting, by a feature vector correcting unit, the feature vector by using the correction vector;
• and
• performing, by a pattern recognition unit, recognition of the input recognition by using the feature vector corrected.