• Documents
  • Authors
  • Tables
  • Log in
  • Sign up
  • MetaCart
  • DMCA
  • Donate

CiteSeerX logo

Tools

Sorted by:
Try your query at:
Semantic Scholar Scholar Academic
Google Bing DBLP
Results 1 - 10 of 47,761
Next 10 →

Table 1: Comparisons between the geometric features based TAN [3] and our LBP-based template matching.

in Conditional Mutual Information Based Boosting for Facial Expression Recognition
by Caifeng Shan, Shaogang Gong, Peter W. Mcowan 2005
"... In PAGE 7: ...2% was obtained by them using Tree-Augmented-Naive Bayes (TAN) classifiers. Comparison in Table1 illustrates that our simple template matching using LBP outperforms geometric features based TAN classifier. The experiments demonstrated that the low-cost LBP features are discriminative for facial expression recogniton.... ..."
Cited by 3

Table 5: The average cost of LBPs for systems with IDD and = 0:8, i = 0:1. In parentheses under LBP-Sum and LBP-1 [respectively, LBP-G, LBP-S] are values of r [resp., ( ; r), ( k; r)]. Values for k 3 are based upon simulation estimates.

in Production-Inventory Systems with Imperfect Advance Demand Information and Updating
by Saif Benjaafar, William L. Cooper, Setareh Mardan 2006
"... In PAGE 38: ... However, LBP-S is simpler because it requires storage of just two parameters. In Table5 , we compare the performance of di erent versions of the LBP heuristic when the number of stages is varied. A common feature of the versions we consider is simplicity of imple- mentation.... ..."

Table 7.4. Average results of the evaluation of the reduction algorithm on the LBP features. The classi er was trained using the one-against-one multi-class algorithm, various kernel types and C = 100.

in ANDRZEJ PRONOBIS Under the Supervision of Barbara Caputo
by Instytut Informatyki, Wydział Automatyki, Elektroniki I Informatyki, Politechnika Śląska Gliwice 2005

Table 7.6. Average results of the evaluation of the reduction algorithm on the LBP features. The classi er was trained using the one-against-all multi-class algorithm, various kernel types and C = 100.

in ANDRZEJ PRONOBIS Under the Supervision of Barbara Caputo
by Instytut Informatyki, Wydział Automatyki, Elektroniki I Informatyki, Politechnika Śląska Gliwice 2005

TABLE II THE RECOGNITION RATES OF THE LBP AND COMPARISON ALGORITHMS.

in unknown title
by unknown authors 2006
Cited by 10

Table 1: Feature distributions across categories

in Category Specific Semantic Deficits In Focal And Widespread Brain Damage: A Computational Account
by Joseph T. Devlin, Laura M. Gonnerman, Elaine S. Andersen, Mark S. Seidenberg 1998
"... In PAGE 7: ... Each word was represented by both perceptual and functional properties. Table1 shows the distribution of features by categories. Overall the model had a 1.... ..."
Cited by 4

Table 3. Distribution of entities and features in data

in Mining Comparative Sentences and Relations
by unknown authors
"... In PAGE 5: ... Different types of sentences Only 4% of the sentences had multiple comparisons in one sentence. The distribution of entities and features in the data set is given in Table3 . Nouns include different types.... ..."

Table 1 Feature Distributions Across Categories

in Category Specific Semantic Deficits in Focal and Widespread Brain Damage: A Computational Account
by Joseph Devlin, Laura M. Gonnerman, Elaine S. Andersen, Mark S. Seidenberg 1998
"... In PAGE 6: ...e.g., found-in-Africa), were not included. Each word was represented by both perceptual and func- tional properties. Table1 shows the distribution of features by categories. Overall the model had a 1.... ..."
Cited by 4

Table 5: Distribution of IM Feature Usage by Students

in Adoption of Instant Messaging Technologies by University Students Abstract
by Farhad Daneshgar, Aybuke Aurum, Sharat Potukuchi 2007
"... In PAGE 6: ... These two features were selected, as it was found that 85% of the respondents were MSN users, and these features were useful tools for coordination and scheduling as well as collaboration. Table5 shows the relative percentage of users that indicated, they were users of each of the features selected. The main features used by the majority of respondents were the file transfer and emoticons , both with 87.... ..."

Table 6. Slot distribution with different feature strings Feature Strings

in Pseudo-Anchor Text Extraction for Vertical Search
by Shuming Shi, Fei Xing, Mingjie Zhu, Zaiqing Nie, Ji-rong Wen
Next 10 →
Results 1 - 10 of 47,761
Powered by: Apache Solr
  • About CiteSeerX
  • Submit and Index Documents
  • Privacy Policy
  • Help
  • Data
  • Source
  • Contact Us

Developed at and hosted by The College of Information Sciences and Technology

© 2007-2019 The Pennsylvania State University