The Experts below are selected from a list of 39 Experts worldwide ranked by ideXlab platform
Andrey V. Savchenko - One of the best experts on this subject based on the ideXlab platform.
-
SACI - Granular Computing and Sequential Analysis of Deep Embeddings in Fast Still-to-Video Face Recognition
2018 IEEE 12th International Symposium on Applied Computational Intelligence and Informatics (SACI), 2018Co-Authors: Andrey V. SavchenkoAbstract:This paper is focused on still-to-video face recognition with large number of subjects based on computation of distances between high-dimensional embeddings extracted using deep convolution neural networks. We propose to utilize granular structures and sequentially process granular representations of all frames of the input video. The coarse-grained granules include only low number of the first principal components of deep embeddings. The representation of each frame at finer granularity levels is matched with the representations of photos of only those individuals, for whom the decision at previous levels was reliable. The reliability is checked by thresholding the ratio of distance between Reference Instance and input frame to the minimal distance. As a result, the photos of all unreliable individuals are not examined anymore for a particular frame at the next levels with finer granularity. Decisions for all frames are united into a candidate set of identities, and the maximal a-posterior final decision is chosen. The experimental study with the LFW, YTF and IJB-A datasets and the state-of-the-art deep embeddings demonstrated that the proposed approach is 2–10 times faster than conventional methods.
-
Maximum-likelihood approximate nearest neighbor method in real-time image recognition
Pattern Recognition, 2017Co-Authors: Andrey V. SavchenkoAbstract:Abstract An exhaustive search of all classes in pattern recognition methods cannot be implemented in real-time, if the database contains a large number of classes. In this paper we introduce a novel probabilistic approximate nearest-neighbor (NN) method. Despite the most of known fast approximate NN algorithms, our method is not heuristic. The joint probabilistic densities (likelihoods) of the distances to previously checked Reference objects are estimated for each class. The next Reference Instance is selected from the class with the maximal likelihood. To deal with the quadratic memory requirement of this approach, we propose its modification, which processes the distances from all Instances to a small set of pivots chosen with the farthest-first traversal. Experimental study in face recognition with the histograms of oriented gradients and the deep neural network-based image features shows that the proposed method is much faster than the known approximate NN algorithms for medium databases.
-
Approximate Nearest Neighbor Search in Intelligent Classification Systems
SpringerBriefs in Optimization, 2016Co-Authors: Andrey V. SavchenkoAbstract:This chapter deals with the problem of insufficient performance of the nearest neighbor-based classification with the medium-sized database (thousands of classes). The key issue of widely applied approximate nearest neighbor algorithms is their heuristic nature. On the contrary, we introduce here a probabilistic approximate NN method by using the asymptotic properties of the classifiers with the segment homogeneity testing from Chap. 2. The joint probabilistic density of the distances to the previously checked Reference objects is estimated for each class at every step. The next Reference Instance to check is selected from the class with the maximal likelihood. Experimental results in image recognition prove that this maximal likelihood search is much more effective for the medium-sized databases, than the brute force and the known approximate nearest neighbor methods.
Jun Pang - One of the best experts on this subject based on the ideXlab platform.
-
ICCD - A novel approach to parameterized verification of cache coherence protocols
2016 IEEE 34th International Conference on Computer Design (ICCD), 2016Co-Authors: Yongjian Li, Yi Lv, Kaiqiang Duan, Jun PangAbstract:Parameterized verification of parameterized protocols like cache coherence protocols is an important but hard problem. Our tool paraVerifier handles this hard problem in a unified framework: (1) it automatically discovers auxiliary invariants and the corresponding causal relations from a small Reference Instance of the verified protocol; (2) the above invariants and causal relation information are automatically generalized into a parameterized form to construct a parameterized formal proof in a theorem prover (e.g., Isabelle). Our method is successfully applied to typical benchmarks including snooping and directory cache coherence protocol benchmarks. The correctness of these protocols is guaranteed by a formal and readable proof which is automatically generated. The notoriously hard FLASH protocol, which is at an industrial scale, is also verified.
-
ATVA - paraVerifier: An automatic framework for proving parameterized cache coherence protocols
Automated Technology for Verification and Analysis, 2015Co-Authors: Yongjian Li, Jun Pang, Yi Lv, Kaiqiang DuanAbstract:Parameterized verification of cache coherence protocols is an important but challenging research problem. We present in this paper our automatic framework paraVerifier to handle this problem: (1) it first discovers auxiliary invariants and thecorresponding causal relations between invariants and protocol rules from a small Reference Instance of the verified protocol; (2) the discovered invariants and causal relations can then be generalized into their parameterized form to automatically construct a formal proof to establish the correctness of the protocol. paraVerifier has been successfully applied to a number of benchmarks.
Kaiqiang Duan - One of the best experts on this subject based on the ideXlab platform.
-
ICCD - A novel approach to parameterized verification of cache coherence protocols
2016 IEEE 34th International Conference on Computer Design (ICCD), 2016Co-Authors: Yongjian Li, Yi Lv, Kaiqiang Duan, Jun PangAbstract:Parameterized verification of parameterized protocols like cache coherence protocols is an important but hard problem. Our tool paraVerifier handles this hard problem in a unified framework: (1) it automatically discovers auxiliary invariants and the corresponding causal relations from a small Reference Instance of the verified protocol; (2) the above invariants and causal relation information are automatically generalized into a parameterized form to construct a parameterized formal proof in a theorem prover (e.g., Isabelle). Our method is successfully applied to typical benchmarks including snooping and directory cache coherence protocol benchmarks. The correctness of these protocols is guaranteed by a formal and readable proof which is automatically generated. The notoriously hard FLASH protocol, which is at an industrial scale, is also verified.
-
ATVA - paraVerifier: An automatic framework for proving parameterized cache coherence protocols
Automated Technology for Verification and Analysis, 2015Co-Authors: Yongjian Li, Jun Pang, Yi Lv, Kaiqiang DuanAbstract:Parameterized verification of cache coherence protocols is an important but challenging research problem. We present in this paper our automatic framework paraVerifier to handle this problem: (1) it first discovers auxiliary invariants and thecorresponding causal relations between invariants and protocol rules from a small Reference Instance of the verified protocol; (2) the discovered invariants and causal relations can then be generalized into their parameterized form to automatically construct a formal proof to establish the correctness of the protocol. paraVerifier has been successfully applied to a number of benchmarks.
Yongjian Li - One of the best experts on this subject based on the ideXlab platform.
-
ICCD - A novel approach to parameterized verification of cache coherence protocols
2016 IEEE 34th International Conference on Computer Design (ICCD), 2016Co-Authors: Yongjian Li, Yi Lv, Kaiqiang Duan, Jun PangAbstract:Parameterized verification of parameterized protocols like cache coherence protocols is an important but hard problem. Our tool paraVerifier handles this hard problem in a unified framework: (1) it automatically discovers auxiliary invariants and the corresponding causal relations from a small Reference Instance of the verified protocol; (2) the above invariants and causal relation information are automatically generalized into a parameterized form to construct a parameterized formal proof in a theorem prover (e.g., Isabelle). Our method is successfully applied to typical benchmarks including snooping and directory cache coherence protocol benchmarks. The correctness of these protocols is guaranteed by a formal and readable proof which is automatically generated. The notoriously hard FLASH protocol, which is at an industrial scale, is also verified.
-
ATVA - paraVerifier: An automatic framework for proving parameterized cache coherence protocols
Automated Technology for Verification and Analysis, 2015Co-Authors: Yongjian Li, Jun Pang, Yi Lv, Kaiqiang DuanAbstract:Parameterized verification of cache coherence protocols is an important but challenging research problem. We present in this paper our automatic framework paraVerifier to handle this problem: (1) it first discovers auxiliary invariants and thecorresponding causal relations between invariants and protocol rules from a small Reference Instance of the verified protocol; (2) the discovered invariants and causal relations can then be generalized into their parameterized form to automatically construct a formal proof to establish the correctness of the protocol. paraVerifier has been successfully applied to a number of benchmarks.
Matthias F. Stallmann - One of the best experts on this subject based on the ideXlab platform.
-
On SAT Instance classes and a method for reliable performance experiments with SAT solvers
Annals of Mathematics and Artificial Intelligence, 2005Co-Authors: Franc Brglez, Xiao Yu Li, Matthias F. StallmannAbstract:A recent series of experiments with a group of state-of-the-art SAT solvers and several well-defined classes of problem Instances reports statistically significant performance variability for the solvers. A systematic analysis of the observed performance data, all openly archived on the Web, reveals distributions which we classify into three broad categories: (1) readily characterized with a simple χ^2-test, (2) requiring more in-depth analysis by a statistician, (3) incomplete, due to time-out limit reached by specific solvers. The first category includes two well-known distributions: normal and exponential; we use simple first-order criteria to decide the second category and label the distributions as near-normal, near-exponential and heavy-tail. We expect that good models for some if not most of these may be found with parameters that fit either generalized gamma, Weibull, or Pareto distributions. Our experiments show that most SAT solvers exhibit either normal or exponential distribution of execution time ( runtime ) on many equivalence classes of problem Instances. This finding suggests that the basic mathematical framework for these experiments may well be the same as the one used to test the reliability or lifetime of hardware components such as lightbulbs, A/C units, etc. A batch of N replicated hardware components represents an equivalence class of N problem Instances in SAT, a controlled operating environment A represents a SAT solver A , and the survival function $$\mathcal{R}^A \left( x \right)$$ (where x represents the lifetime ) is the complement of the solvability function $$\mathcal{S}^A \left( x \right) = 1--\mathcal{R}^A \left( x \right)$$ where x may represent runtime, implications, backtracks , etc. As demonstrated in the paper, a set of unrelated benchmarks or randomly generated SAT Instances available today cannot measure the performance of SAT solvers reliably — there is no control on their ‘hardness’. However, equivalence class Instances as defined in this paper are, in effect, replicated Instances of a specific Reference Instance. The proposed method not only provides a common platform for a systematic study and a reliable improvement of deterministic and stochastic SAT solvers alike but also supports the introduction and validation of new problem Instance classes.
-
On SAT Instance classes and a method for reliable performance experiments with SAT solvers
Annals of Mathematics and Artificial Intelligence, 2005Co-Authors: Franc Brglez, Xiao Yu Li, Matthias F. StallmannAbstract:A recent series of experiments with a group of state-of-the-art SAT solvers and several well-defined classes of problem Instances reports statistically significant performance variability for the solvers. A systematic analysis of the observed performance data, all openly archived on the Web, reveals distributions which we classify into three broad categories: (1) readily characterized with a simple χ2-test, (2) requiring more in-depth analysis by a statistician, (3) incomplete, due to time-out limit reached by specific solvers. The first category includes two well-known distributions: normal and exponential; we use simple first-order criteria to decide the second category and label the distributions as near-normal, near-exponential and heavy-tail. We expect that good models for some if not most of these may be found with parameters that fit either generalized gamma, Weibull, or Pareto distributions. Our experiments show that most SAT solvers exhibit either normal or exponential distribution of execution time (runtime) on many equivalence classes of problem Instances. This finding suggests that the basic mathematical framework for these experiments may well be the same as the one used to test the reliability or lifetime of hardware components such as lightbulbs, A/C units, etc. A batch of N replicated hardware components represents an equivalence class of N problem Instances in SAT, a controlled operating environment A represents a SAT solver A, and the survival function RA(x) (where x represents the lifetime) is the complement of the solvability function SA(x)e1−RA(x) where x may represent runtime, implications, backtracks, etc. As demonstrated in the paper, a set of unrelated benchmarks or randomly generated SAT Instances available today cannot measure the performance of SAT solvers reliably – there is no control on their ‘hardness’. However, equivalence class Instances as defined in this paper are, in effect, replicated Instances of a specific Reference Instance. The proposed method not only provides a common platform for a systematic study and a reliable improvement of deterministic and stochastic SAT solvers alike but also supports the introduction and validation of new problem Instance classes.