Method of recognition of human motion, vector sequences and speech
First Claim
1. A method of recognizing and classifying an input record of human movements as a member of a class of records of model human activity Aj in a collection of classes of records of model human activities {Aj}, comprising:
- (a) predetermining and recording said collection of classes of records of model human activities {Aj};
(b) obtaining from each said class of records of model human activity Aj at least one model sample;
(c) representing each said model sample by a model vector mrj;
(d) representing each said class of records of model human activity Aj by a model vectors sequence Mj=(mlj . . . mrj . . . mqj), wherein each said model vector mrj in said model vectors sequence Mj corresponds to said model sample and wherein the subscript q denotes the total number of model vectors in said model vectors sequence Mj, and wherein the subscript r denotes the location of said model vector mrj within said model vectors sequence Mj and wherein the subscript j denotes the serial number of said model vectors sequence Mj and also denotes the same serial number of the corresponding said class of records of model human activity Aj;
(e) representing said collection of classes of records of model human activities {Aj} by a collection of corresponding model vectors sequences {Mj};
(f) constructing a table with multidimensional addressing of table bins;
(g) predetermining a distance threshold value;
(h) storing each of said model vector mrj in a table bin whose address has a multidimensional value which has the smallest multidimensional distance to a multidimensional value of said model vector mrj;
(i) obtaining from said input record of human movements at least one input sample;
(j) representing each of said input sample by an input vector tnk;
(k) representing said input record of human movements by an input vectors sequence Tk =(tlk . . . tnk . . . tpk), wherein the subscript p denotes the total number of the input vectors in said input vectors sequence Tk, and wherein the subscript n denotes the location of said input vector tnk within said input vectors sequence Tk, and wherein the subscript k denotes a serial number of said input vectors sequence;
(l) for each said input vector tnk selecting all the table bins with multidimensional address values which have multidimensional distances to the multidimensional value of said input vector tnk, which are below said distance threshold value;
(m) for each selected said table bin, retrieving all said model vectors mrj that were stored in said selected said table bin;
(n) employing a matching algorithm that uses said input vectors sequence Tk and the retrieved said model vectors mrj, to produce a collection of matching scores {Skj}, wherein each matching score Skj denotes the degree of similarity between said input vectors sequence Tk and one said model vectors sequence Mj, which is a member of said collection of model vectors sequences {Mj};
(o) recognizing and classifying said input record of human movements as a member of said class of records of model human activity Aj which is represented by said model vectors sequence Mj with the highest said matching score Skj in said collection of matching scores {Skj};
(p) adding information about said class of records of model human activity Aj that was selected, to contents of said input record of human movements and providing a combined input record of human movements as an output record.
0 Assignments
0 Petitions
Accused Products
Abstract
A method for recognition of an input human motion as being the most similar to one model human motion out of a collection of stored model human motions. In the preferred method, both the input and the model human motions are represented by vector sequences that are derived from samples of angular poses of body parts. The input and model motions are sampled at substantially different rates. A special optimization algorithm that employs sequencing constraints and dynamic programming, is used for finding the optimal input-model matching scores. When only partial body pose information is available, candidate matching vector pairs for the optimization are found by indexing into a set of hash tables, where each table pertains to a sub-set of body parts. The invention also includes methods for recognition of vector sequences and for speech recognition.
37 Citations
20 Claims
-
1. A method of recognizing and classifying an input record of human movements as a member of a class of records of model human activity Aj in a collection of classes of records of model human activities {Aj}, comprising:
-
(a) predetermining and recording said collection of classes of records of model human activities {Aj}; (b) obtaining from each said class of records of model human activity Aj at least one model sample; (c) representing each said model sample by a model vector mrj; (d) representing each said class of records of model human activity Aj by a model vectors sequence Mj=(mlj . . . mrj . . . mqj), wherein each said model vector mrj in said model vectors sequence Mj corresponds to said model sample and wherein the subscript q denotes the total number of model vectors in said model vectors sequence Mj, and wherein the subscript r denotes the location of said model vector mrj within said model vectors sequence Mj and wherein the subscript j denotes the serial number of said model vectors sequence Mj and also denotes the same serial number of the corresponding said class of records of model human activity Aj; (e) representing said collection of classes of records of model human activities {Aj} by a collection of corresponding model vectors sequences {Mj}; (f) constructing a table with multidimensional addressing of table bins; (g) predetermining a distance threshold value; (h) storing each of said model vector mrj in a table bin whose address has a multidimensional value which has the smallest multidimensional distance to a multidimensional value of said model vector mrj; (i) obtaining from said input record of human movements at least one input sample; (j) representing each of said input sample by an input vector tnk; (k) representing said input record of human movements by an input vectors sequence Tk =(tlk . . . tnk . . . tpk), wherein the subscript p denotes the total number of the input vectors in said input vectors sequence Tk, and wherein the subscript n denotes the location of said input vector tnk within said input vectors sequence Tk, and wherein the subscript k denotes a serial number of said input vectors sequence; (l) for each said input vector tnk selecting all the table bins with multidimensional address values which have multidimensional distances to the multidimensional value of said input vector tnk, which are below said distance threshold value; (m) for each selected said table bin, retrieving all said model vectors mrj that were stored in said selected said table bin; (n) employing a matching algorithm that uses said input vectors sequence Tk and the retrieved said model vectors mrj, to produce a collection of matching scores {Skj}, wherein each matching score Skj denotes the degree of similarity between said input vectors sequence Tk and one said model vectors sequence Mj, which is a member of said collection of model vectors sequences {Mj}; (o) recognizing and classifying said input record of human movements as a member of said class of records of model human activity Aj which is represented by said model vectors sequence Mj with the highest said matching score Skj in said collection of matching scores {Skj}; (p) adding information about said class of records of model human activity Aj that was selected, to contents of said input record of human movements and providing a combined input record of human movements as an output record. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A method of recognizing and classifying a record of input signals as a member of a class of records of model signals Aj in a collection of classes of records of model signals {Aj}, comprising:
-
(a) predetermining and recording said collection of classes of records of model signals {Aj}; (b) obtaining from each said class of records of model signals Aj at least one model sample; (c) representing each said model sample by a model vector mrj; (d) representing each said class of records of model signals by a model vectors sequence Mj=(mlj . . . mrj . . . mqj), wherein each said model vector mrj in said model vectors sequence Mj corresponds to said model sample and wherein the subscript q denotes the total number of model vectors in said model vectors sequence Mj, and wherein the subscript r denotes the location of said model vector mrj within said model vectors sequence Mj and wherein the subscript j denotes the serial number of said model vectors sequence Mj and also denotes the same serial number of the corresponding said class of records of model signals Aj; (e) representing said collection of classes of records of model signals {Aj} by a collection of corresponding model vectors sequences {Mj}; (f) constructing a table with multidimensional addressing of table bins; (g) predetermining a distance threshold value; (h) storing each of said model vector mrj in a table bin whose address has a multidimensional value which has the smallest multidimensional distance to a multidimensional value of said model vector mrj; (i) obtaining from said record of input signals at least one input sample; (j) representing each of said input sample by an input vector tnk; (k) representing said record of input signals by an input vectors sequence Tk=(tlk . . . tnk . . . tpk), wherein the subscript p denotes the total number of the input vectors in said input vectors sequence, and wherein the subscript n denotes the location of said input vector tnk within said input vectors sequence Tk, and wherein the subscript k denotes a serial number of said input vectors sequence; (l) for each said input vector tnk selecting all the table bins with multidimensional address values which have multidimensional distances to the multidimensional value of said input vector tnk, which are below said distance threshold value; (m) for each selected said table bin, retrieving all said model vectors mrj that were stored in said selected said table bin; (n) employing a matching algorithm that uses the retrieved said model vectors mrj and said input vectors sequence Tk, to produce a collection of matching scores {Skj}, wherein each matching score Skj denotes the degree of similarity between said input vectors sequence Tk and one said model vectors sequence Mj, which is a member of said collection of model vectors sequences {Mj}; (o) recognizing and classifying said record of input signals as a member of said class of records of model signals Aj which is represented by said model vectors sequence Mj that has the highest said matching score Skj in said collection of matching scores {Skj}; (p) adding information about said class of records of model signals Aj that was selected, to contents of said record of input signals and providing a combined record of input signals as an output record. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
Specification