### TABLE I STATISTICS OF RIGID HEAD MOTION

### Table 1: Comparison between the geometry for two views of 1 rigid body motion and that for n rigid body motions.

"... In PAGE 4: ... 5.5 Summary of multibody geometry Table1 summarizes our theoretical development with a comparison of the geometric entities associated with two views of 1 rigid body motion and two views of n rigid body motions. Polynomial factorization Segmentation Epipolar transfer Veronese map PSfrag replacements x1 2 R3 x2 2 R3 n(x1) 2 RMn [x;y;z]T 7! [:::;xn1yn2zn3;:::]T F n(x1) 2 RMn xT 2 lk = 0 (x1;x2) 2 ith motion eT i lk = 0; or xT 2 Fix1 = 0 n(x)TF n(x1) = (xTl1) (xTln) (eT 1 l) (eT nl) = 0 lk 2 R3 fe1;:::;en 2 R3g fl1;:::;ln 2 R3g Figure 1: Transformation diagram associated with the segmentation of an image pair (x1;x2) in the presence of n motions.... ..."

### Table I. Comparison between the geometry for two views of 1 rigid body motion and that for n rigid body motions.

2004

Cited by 25

### Table 1. Comparison between the geometry for two views of 1 rigid-body motion and the geometry of n rigid-body motions.

2004

Cited by 25

### Table 1. Comparison between the geometry for two views of 1 rigid-body motion and the geometry of n rigid-body motions.

2004

Cited by 25

### Table I. Comparison between the geometry for two views of 1 rigid-body motion and the geometry of n rigid-body motions.

2004

Cited by 25

### Table 1: The simulated 3D rigid motions and the retinal locations of their corresponding epipoles.

1999

Cited by 11

### Table 1: Comparison of estimation errors for rigid motion between CCRE, MI and normalized MI.

"... In PAGE 4: ... These are normally dis- tributed around the values of (0 ; 5pixel; 5pixel), with standard deviations of ( 8 , 3pixel 3pixel) for rotation and translation in x and y respectively. Table1 shows the statis- tics of errors resulting from the 3 different methods. In each cell, the leftmost value is the rotation angle (in de- grees), while the right two values show the translations in x and y directions.... ..."

### Table 3. Results using PCA on the rigid head motion sequences. Compare to middle sec-

"... In PAGE 7: ... These eigenvectors were then used as a basis on which flows were projected, the results of which were used to train HMMs, as before. Table3 shows recognition re- sults for using a subset of the principal components, starting from the third, and can be directly compared to the results in the middle section of Table 2. Similar results were ob- tained in comparison to the other two sections of Table 2.... ..."

### Table 3. Results using PCA on the rigid head motion sequences. Compare to middle sec-

"... In PAGE 7: ... These eigenvectors were then used as a basis on which flows were projected, the results of which were used to train HMMs, as before. Table3 shows recognition re- sults for using a subset of the principal components, starting from the third, and can be directly compared to the results in the middle section of Table 2. Similar results were ob- tained in comparison to the other two sections of Table 2.... ..."