### Table 1: The accuracy of the calculated poses

1994

"... In PAGE 5: ... Both objects were correctly localized although some of the edges were occluded and some were simply not detected. Table1 shows some re- sults presenting the precision of the calculated pose of the ob- ject which can be seen in the right parts of both images shown in Fig. 3.... ..."

Cited by 3

### Table 1: Simulations on observing the poses and motions of random polygons being pushed

1998

"... In PAGE 32: ... observation. 20 Table1 summarizes the results with known contact edges. There are four groups of data, each representing a di#0Berent combination of contact mode and disk acceleration.... In PAGE 32: ... Since in every test the estimated contact point, given by s, was randomly chosen on the contact edge, it could be far from the real contact point. Yet, the results in Table1 seem to suggest that the local GHO observer has #5Cglobalness quot;, at least within one edge. We also observed that the disk contact estimate ~ u and the angular velocity estimate ~ ! always converged very fast, and the tangential velocity estimate ~ v T almost always converged.... ..."

### Table 4: Spatio-temporal queries posed on Precipitation

2003

"... In PAGE 8: ... Drill-down Queries. Our implementation considers four types of queries (shown in Table4 ) that involve different extents of spatio- temporal processing that evaluate both advantages and limitations of wavelet compression. The GlobalYearlyEdge and LocalYear- lyMean queries explore features for which wavelet processing is typically well suited.... In PAGE 9: ... Our goal in this section is to demonstrate the search features of the system and prove our claim that multi-resolution storage can be useful for a broad variety of queries. To evaluate performance, each of the queries shown in Table4 was posed over the dataset. For yearly queries (GlobalYearlyEdge and and GlobalYearlyMax), there were 45 instances each, since there are 45 years of data.... In PAGE 10: ... Figure 6.2 shows the variation of query quality for queries defined in Table4 for different levels of drill-down. Performance for LocalYearlyMean, GlobalYearlyMax and Local- DailyMax queries are very similar, as shown in Figure 6.... In PAGE 10: ... Evaluating Training using Limited Information In our evaluation, we use a training period of two epochs of data (10% of total deployment time) to predict the query accuracy for the entire dataset. Summaries are constructed over the training set, and all queries in Table4 are posed over these summaries. Ideally, the error obtained from the training set would mirror error seen by the omniscient scheme.... In PAGE 11: ... The first column in Table 7 shows the difference between the per- formance of training and the optimal schemes. These results are ag- gregate results over a range of storage sizes (0 - 100KB) and query types (shown in Table4 ). Training performs exceedingly well, and in fact is on average less than 1% worse than the optimal solution.... ..."

Cited by 53

### Table 4: Spatio-temporal queries posed on Precipitation

2003

"... In PAGE 8: ... Drill-down Queries. Our implementation considers four types of queries (shown in Table4 ) that involve different extents of spatio- temporal processing that evaluate both advantages and limitations of wavelet compression. The GlobalYearlyEdge and LocalYear- lyMean queries explore features for which wavelet processing is typically well suited.... In PAGE 9: ... Our goal in this section is to demonstrate the search features of the system and prove our claim that multi-resolution storage can be useful for a broad variety of queries. To evaluate performance, each of the queries shown in Table4 was posed over the dataset. For yearly queries (GlobalYearlyEdge and and GlobalYearlyMax), there were 45 instances each, since there are 45 years of data.... In PAGE 9: ... Figure 6.2 shows the variation of query quality for queries defined in Table4 for different levels of drill-down. Performance for LocalYearlyMean, GlobalYearlyMax and Local- DailyMax queries are very similar, as shown in Figure 6.... In PAGE 10: ... Evaluating Training using Limited Information In our evaluation, we use a training period of two epochs of data (10% of total deployment time) to predict the query accuracy for the entire dataset. Summaries are constructed over the training set, and all queries in Table4 are posed over these summaries. Ideally, the error obtained from the training set would mirror error seen by the omniscient scheme.... In PAGE 11: ... The first column in Table 7 shows the difference between the per- formance of training and the optimal schemes. These results are ag- gregate results over a range of storage sizes (0 - 100KB) and query types (shown in Table4 ). Training performs exceedingly well, and in fact is on average less than 1% worse than the optimal solution.... ..."

Cited by 53

### Table 2: Simulations of the GHO observers #2832#29 and #2843#29 on #0Cnding the poses and motions

1998

"... In PAGE 33: ... The observer fails if all estimates have gone out of their ranges in the #0Crst period, or the obtained contact estimate, including the edge and the location on the edge, is incorrect. Table2 shows the test results with unknown contact edges. A high percentage of the failures reported in the table were due to incorrect contact edges.... ..."

### Table 1. Influence of small changes of the pose parameters on the observed photopolarimetric, geometric, and depth cues.

"... In PAGE 5: ... We found that the influence on the observed intensity, polarisation, edge, and depth cues is different for small vari- ations of each pose parameter (cf. Table1 ). For example, a slight lateral translation has a strong influence on the edges in the image but may leave the observed intensity and po- larisation angle largely unchanged.... ..."

### Table 3: Total number of Newton steps during a simulation on a mesh with 356701 edges decom- posed into 16 subdomains. The varying parameters are the preconditioner and the scaling strategy. means that the nonlinear scheme does not converge.

"... In PAGE 10: ... The preconditioners considered are block Jacobi denoted by MbJ and the additive Schwarz preconditioner denoted by MAS. In Table3 we display the total number of Newton steps (and therefore the total number of linear systems solved) required to obtain the stationary state of the simulation. We consider a diagonal scaling on the original systems or a diagonal scaling on the corresponding Schur complements.... In PAGE 11: ...osed into 16 subdomains. The varying parameters are the preconditioner and the scaling strategy. means that the nonlinear scheme does not converge. Table3 shows that a combination of a preconditioner for S and a scaling on S is needed to ensure the convergence of the nonlinear scheme. Actually, the Schur complement matrices are e ectively very badly scaled with values varying with more than 30 orders of magnitude.... ..."

### Table 4. CPSNR performance (in dB) of LSAI, MML, and the pro- posed OCP methods

"... In PAGE 3: ... For fairness, we adopt an orthogonal wavelet transform for all wavelet-based tested approaches. As shown in Table4 4, the proposed approach outperforms all others on every test image. OCP denoising approach performs especially well near the edges.... ..."

### Table 1: Confusion matrix for the recognition results of 10 unoccluded views of each object.

"... In PAGE 4: ... The resolution of these views was 400 faces per view. Table1 shows the confusion matrix for our recognition results. Notice that there is only one false-positive i.... In PAGE 5: ... 5. Note that the recognition rate is 100% at zero occlusion because these tests were performed on a subsample of unoccluded views which were correctly recognized in Table1 . Fig.... In PAGE 5: ... After the addition of Gaussian noise with = 2:0cm. sample of noiseless views which were correctly recognized in Table1 . The recognition rate was 100% up to a noise with standard deviation of 2.... ..."

### Table 1 Poses used for synthetic images. Pose

1996

"... In PAGE 5: ... Three different component models were used: one tibial implant and two femoral implants. Each component was rendered in 8 different poses ( Table1 ) for a total of 24 test images. Table 1 Poses used for synthetic images.... ..."

Cited by 2