### TABLE 1: CONSTRAINTS FOR CERTAIN PROCESSES

Cited by 1

### Table 5.3: Number of consistency checks performed by various backtracking algo- rithms on certain constraint satisfaction problems.

### Table 12: Resulting discriminant accuracy using MAV 1 and di erent used smoothing parameters and mapping schemes with the TIMIT database. The used plane index for the optimal accuracy is given for each representation which translates to a certain time constraints . In this experiment we warped the test data into the template before comparison.

1999

Cited by 8

### Table 9: Resulting discriminant accuracy using MAV and di erent smoothing parameters . The mapping scheme for the ISOLET database is based on the assumption that all diphones consists of the same number of frames because of the missing phone labels. The used plane index for the optimal accuracy is given for each representation which translates to a certain time constraints .

1999

Cited by 8

### Table 10: Resulting discriminant accuracy using the Principal Curve model MPC and di erent smoothing parameters . The mapping scheme for the ISOLET database is based on the assumption that all diphones consists of the same number of frames because of the missing phone labels. The used plane index for the optimal accuracy is given for each representation which translates to a certain time constraints .

1999

"... In PAGE 25: ... With the best results reaching an average accuracy of 81.11%, using our principal curve model MPC with an appropriate smoothing factor (see Table10 ) , we obtain similar results to a baseline HMM system using one mixture and a diagonal covariance matrix on a BTL E-SET giving 84.5% accuracy for this task.... ..."

Cited by 8

### Table 11: Resulting discriminant accuracy using the Generative Topographic Mapping model MGTM and di erent smoothing parameters . The mapping scheme for the ISOLET database is based on the assumption that all diphones consists of the same number of frames because of the missing phone labels. The used plane index for the optimal accuracy is given for each representation which translates to a certain time constraints .

1999

Cited by 8

### Table13: Resulting discriminant accuracy using MAV 2 and di erent used smoothing parameters and mapping schemes with the TIMIT database. The used plane index for the optimal accuracy is given for each representation which translates to a certain time constraints . In this experiment we adjusted the test trajectory to the template using Schwartz inelastic region idea.

1999

Cited by 8

### Table 14: Resulting discriminant accuracy using the Principal Curve model MPC and di erent used smoothing parameters and mapping schemes with the TIMIT database. The used plane index for the optimal accuracy is given for each representation which translates to a certain time constraints . In this experiment we adjusted the test trajectory to the size of the template using time warping.

1999

Cited by 8

### Table 15: Resulting discriminant accuracy using the Generative Topographic Mapping model MGTM and di erent used smoothing parameters and mapping schemes with the TIMIT database. The used plane index for the optimal accuracy is given for each representation which translates to a certain time constraints . In this experiment we adjusted the test trajectory to the size of the template using time warping.

1999

Cited by 8