### Table 1. Overview of Fusion Techniques

"... In PAGE 2: ... This work has produced numerous techniques, which can be decomposed into five categories: classifier selection, combination of classifier outputs, sampling of classifier training data, manipulation of classifier outputs, and classifier feature selection. Classifier fusion techniques are categorized in Table1 , and a brief explanation of each technique follows. Classifier Selection Classifier selection endeavors to choose the best classifier for a given task.... ..."

### Table 1: Resistance of a single contact to the backplane using the three techniques

1995

"... In PAGE 59: ...ubstrate. The computation time for the numerical and the DCT based techniques are compared. The computation time in the numerical scheme is small for a small number of grid points, but becomes large as the number of points is increased. It can be seen from Table1 that the accuracy of the result in the numerical technique is poor for a small number of grid points. Table 1 also lists the computation time if the [P] matrix is computed explicitly without using the DCT.... In PAGE 59: ... It can be seen from Table 1 that the accuracy of the result in the numerical technique is poor for a small number of grid points. Table1 also lists the computation time if the [P] matrix is computed explicitly without using the DCT. As can be seen the computation time is large.... ..."

### Table 1: Usability of present day eye-gaze techniques. The information has been gathered from numerous places (Scott amp; Findlay, 1993; Baluja amp; Pomerleau, 1994; Wooding, 1995, among others).

1995

"... In PAGE 46: ... Errors in the program should be corrected. Table1 0: Important problems with the EyeCatcher... In PAGE 57: ...Advantages Disadvantages Keyboard Enables free composition of in- put, precise spelling and precise magnitude input (one can spec- ify (42,17) instead of pointing at the approximate spot) Input rate is low and constrained to character-based descriptions Mouse Enables semi-precise positioning and input of position or shape- based descriptions, and includes a take/drop function (mouse button) Input is constrained to pointing and clicking on menus (non-free compositon) Eye-gaze Input rate is high and shows the point of interest The eyes are continuously oper- ated, do not have a take/drop function and input is constrained to pointing and menu-selection (non-free composition) Speech Enables free composition of input and does not require syntactic knowledge of keyboards Input is ambiguous (homonyms cannot be distinguished) and can disturb the surroundings Table1 1: Communication advantages and disadvantages of the di erent modes of communication from human to computer We believe that the ideal eye-gaze application should use several dimen- sions of the eye-tracking data. As described in section 3 on page 12, the nal eye-movements are caused by many di erent cognitive processes, so di erent aspects of the user apos;s cognitive state are exposed by di erent fea- tures of the gaze pattern.... In PAGE 65: ...Voluntary Time context Change in pupil size ? Short Altered blinking rate ? Short Head orientation + Short Sudden head movements ? Short Gestures + Short Change in heartbeat volume ? Short Uneasiness / nervous movements + Short Voice characteristica + Short Eye contact avoidance + Short Alterations in pulse rate ? Long Breathing rate ? Long Body temperature ? Long Sweat rate ? Long Table1 2: A ection measures within a short time interval and those which demand long time interpretation sampling intervals (cf. gure 12 on the next page).... ..."

Cited by 14

### Table 1: Summary of the computational requirements A bene t of using the new technique to deal with the ladder network can be seen on numerical exam- ples using some typical values for n given in [9]. In

1994

Cited by 3

### Table 9.5: Comparison: direct calculation versus the nested iteration technique in the second numerical experiment for the L shape problem. Considered is one time-step with t0 = 0 and t1 = 1 and six uniform triangulations with 6, 24, 94, 384, 1536 and 6114 elements.

2001

### Table 2: Numerical results for the domain transformation technique for solving the 2D Laplace equation with prescribed surface shapes 1, 2 and 3. =1.0e-04 =1.0e-06 =1.0e-08 =1.0e-10 N

"... In PAGE 10: ... Gaussian elimination is used as the equation solver, and linear elements (triangles with three nodes) are used for the discretization. The nu- merical results are shown in Table2 , where nite element solutions are denoted by b apos;. Clearly, second order convergence is obtained for all test problems.... ..."

### Table 2: Numerical Data

"... In PAGE 5: ... As a result of this variance reduction technique, the number of replica- tions to run in order to get an estimate with a given relative error is, on average, at least 60% less. 5 NUMERICAL EXAMPLE The numerical data assumed to test the algorithm are re- ported in Table2 . In the following example we did not con- sider any historical data, because it is not a real case study.... ..."

### Table 2: Numerical results for the domain transformation technique for solving the 2D Laplace equation with prescribed surface shapes 1, 2 and 3. 3.2.2 Numerical results for the PCG method Now we want to study the number of CG-iterations needed to solve the linear system associated with the two-dimensional test problem described above. A vector 0 with

"... In PAGE 11: ... Gaussian elimination is used as the equation solver, and linear elements (triangles with three nodes) are used for the discretization. The numerical results are shown in Table2 , where nite element solutions are denoted by b apos;. Clearly, second order convergence is obtained for all test problems.... ..."

### Table 1: Numerical results for different cylinder approximations. row 1: number of faces of the cylinder, cols1-5:errors for each technique of col 6

Cited by 1