Abstract: The present disclosure provides a method for recognition. The method includes inputting a melody and obtaining pitch tracking information of the melody; obtaining beat information of the melody; determining a clarity value according to the pitch tracking information; implementing a first comparison process first to filter a first set of candidate songs from a database and then implementing a second comparison process to filter a second set of candidate songs from the first set of candidate songs if the clarity value is larger than a predetermined threshold; and determining at least one final candidate song from the second set of candidate songs.

Abstract: A route recommendation system is provided. The system has: a data querying apparatus configured to generate query information according to a location; and a data analyzing apparatus, connected to the data querying apparatus, having: a database configured to store at least one candidate data, wherein each candidate data has a candidate location, at least one recommended location corresponding to the candidate location, and a recommended staying time of each recommended location; a data receiving unit configured to receive the query information; and a processing unit configured to generate at least one recommended route according to the query information and the candidate data, and transmit the at least one recommended route to the data querying apparatus, wherein the recommended route comprises a path from the location to the at least one recommended location, and the recommended staying time corresponding to each recommended location.

Abstract: The present disclosure provides a method for recognition. The method includes inputting a melody and obtaining pitch tracking information of the melody; obtaining beat information of the melody; determining a clarity value according to the pitch tracking information; implementing a first comparison process first to filter a first set of candidate songs from a database and then implementing a second comparison process to filter a second set of candidate songs from the first set of candidate songs if the clarity value is larger than a predetermined threshold; and determining at least one final candidate song from the second set of candidate songs.

Abstract: An adjustable hierarchical scoring method is disclosed. A hierarchical scoring structure for speech-sound data is generated. Historical data of speech-sound scoring for the speech-sound data is retrieved from a speech-sound practice database while a professional scoring result for the hierarchical scoring structure of the speech-sound data is being collected. A weight-adjusting operation is performed according to the historical data of speech-sound scoring and the professional scoring result to calculate optimum adjusting weights for each layer of the hierarchical scoring structure. Retrieved speech-sound data of a student is scored using a speech-sound scoring system and a weighted average operation is performed to the score of the speech-sound data of the student according to a corresponding adjusting weight, thereby generating the final score of the speech-sound data of the student.

Abstract: A method of generating and detecting confusing phones/syllables is disclosed. The method includes a generating stage and a detecting stage. The generating stage includes: (a) input a Mandarin utterance; (b) partition the Mandarin utterance into segmented phones/syllables and generate the most likely route in a recognition net via Forced Alignment of Viterbi decoding; (c) compare the segmented phones/syllables with a Mandarin acoustic model; (d) determine whether a confusing phone/syllable exists; (e) add the confusing phone/syllable into the recognition net and repeat step (b), (c), and (d) when the confusing phone/syllable exists; (f) stop and output all generated confusing phones/syllables to a confusing phone/syllable file when a confusing phone/syllable does not exist.

Abstract: A singing voice synthesis system is provided. The storage unit stores at least one tune. The tempo unit provides a set of tempo cues in accordance with a selected tune from the at least one tune. The input unit receives a plurality of original voice signals corresponding to the selected tune. The processing unit processes the original voice signals and generates a synthesized singing voice signal according to the selected tune.

Abstract: An adjustable hierarchical scoring method is disclosed. A hierarchical scoring structure for speech-sound data is generated. Historical data of speech-sound scoring for the speech-sound data is retrieved from a speech-sound practice database while a professional scoring result for the hierarchical scoring structure of the speech-sound data is being collected. A weight-adjusting operation is performed according to the historical data of speech-sound scoring and the professional scoring result to calculate optimum adjusting weights for each layer of the hierarchical scoring structure. Retrieved speech-sound data of a student is scored using a speech-sound scoring system and a weighted average operation is performed to the score of the speech-sound data of the student according to a corresponding adjusting weight, thereby generating the final score of the speech-sound data of the student.

Abstract: A method of generating and detecting confusing phones/syllables is disclosed. The method includes a generating stage and a detecting stage. The generating stage includes: (a) input a Mandarin utterance; (b) partition the Mandarin utterance into segmented phones/syllables and generate the most likely route in a recognition net via Forced Alignment of Viterbi decoding; (c) compare the segmented phones/syllables with a Mandarin acoustic model; (d) determine whether a confusing phone/syllable exists; (e) add the confusing phone/syllable into the recognition net and repeat step (b), (c), and (d) when the confusing phone/syllable exists; (f) stop and output all generated confusing phones/syllables to a confusing phone/syllable file when a confusing phone/syllable does not exist.

Abstract: A neural network uses a fuzzy membership function, the parameters of which are adaptive during the training process, to parameterize the interconnection weights between an (n-1)'th layer and an n'th layer of the network. Each j'th node in each k'th layer of the network except the input layer produces its output value y.sub.k,j according to the function ##EQU1## where N.sub.k-1 is the number of nodes in layer k-1, i indexes the nodes of layer k-1 and all the w.sub.k,i,j are interconnection weights. The interconnection weights to all nodes j in the n'th layer are given by w.sub.n,i,j =w.sub.n,j (i, p.sub.n,j,1, . . . , p.sub.n,j,p.sbsb.n).The apparatus is trained by setting values for at least one of the parameters p.sub.n,j,1, . . . , p.sub.n,j,Pn. Preferably the number of parameters P.sub.n is less than the number of nodes N.sub.n-1 in layer n-1. w.sub.n,j (i,p.sub.n,j,1, . . . , p.sub.n,j,Pn) can be convex in i, and it can be bell-shaped. Sample functions for w.sub.n,j (i, p.sub.n,j,1, . . . , p.sub.