In this study, we first collected ECG data at 32 ksps and 500 sps from 51 patients with an implanted cardiac pacemaker. We also included 16 non paced ECGs with severe motion artifact and static interference. We then a...
详细信息
In this study, we first collected ECG data at 32 ksps and 500 sps from 51 patients with an implanted cardiac pacemaker. We also included 16 non paced ECGs with severe motion artifact and static interference. We then annotated the pacer pulses (locations and classifications) using special viewing software and divided the data into training and testing sets. We developed three digital pacemaker stimulus detection algorithms, one for each of three different sampling rates (500 sps, 8 ksps, and 32 ksps) by using similar detection concepts. After tuning them on the training sets at the corresponding sampling rates, we investigated their performances. Results showed that the detection algorithm based on the 12-lead 500-sps data stream has a good performance (sensitivity or Se=77.51%; positive productive value, or +P=90.32%). The 8-ksps detector has a better performance (Se=95.37; +P=99.77). The digital pacemaker detection algorithm for 32 ksps or higher data provided almost 100% true pacer spike detection (Se=99.51 and +P=100%). In summary, high speed sampling at 32 ksps can provide accurate detection of modern pacers and leads to a significant improvement in detection of pacemaker stimuli compared to the reduced sampling rates used previously.
This paper introduces a new extended average magnitude difference function (EAMDF) for noise robust pitch detection. EAMDF involves in sufficient number of averaging for all lag values compared to the original AMDF, a...
详细信息
This paper introduces a new extended average magnitude difference function (EAMDF) for noise robust pitch detection. EAMDF involves in sufficient number of averaging for all lag values compared to the original AMDF, and thereby eliminates the falling tendency of the AMDF without emphasizing pitch harmonics at higher lags, which is a severe limitation of other existing improvements of the AMDF. A noise robust post processing that explores the contribution of each frequency channel is also presented. Experimental results on Keele pitch database in different noise level shows the superiority of the proposed EAMDF based pitch detection method over other methods based on AMDF.
This paper addresses human pose recognition from video sequences by formulating it as a classification problem. Unlike much previous work we do not make any assumptions on the availability of clean segmentation. The f...
详细信息
This paper addresses human pose recognition from video sequences by formulating it as a classification problem. Unlike much previous work we do not make any assumptions on the availability of clean segmentation. The first step of this work consists in a novel method of aligning the training images using 3D Mocap data. Next we define classes by discretizing a 2D manifold whose two dimensions are camera viewpoint and actions. Our main contribution is a pose detection algorithm based on random forests. A bottom-up approach is followed to build a decision tree by recursively clustering and merging the classes at each level. For each node of the decision tree we build a list of potentially discriminative features using the alignment of training images;in this paper we consider Histograms of Orientated Gradient (HOG). We finally grow an ensemble of trees by randomly sampling one of the selected HOG blocks at each node. Our proposed approach gives promising results with both fixed and moving cameras.
We present Nodeinfo, an unsupervised algorithm for anomaly detection in system logs. We demonstrate Nodeinfo's effectiveness on data from four of the world's most powerful supercomputers: using logs representi...
详细信息
We present Nodeinfo, an unsupervised algorithm for anomaly detection in system logs. We demonstrate Nodeinfo's effectiveness on data from four of the world's most powerful supercomputers: using logs representing over 746 million processor-hours, in which anomalous events called alerts were manually tagged for scoring, we aim to automatically identify the regions of the log containing those alerts. We formalize the alert detection task in these terms, describe how Nodeinfo uses the information entropy of message terms to identify alerts, and present an online version of this algorithm, which is now in production use. This is the first work to investigate alert detection on (several) publicly-available supercomputer system logs, thereby providing a reproducible performance baseline.
The paper presents a low cost skip detection algorithm and its architecture by detecting 4x4-zero block numbers in a macroblock (MB). With this simple detection mechanism and an adaptive threshold; the proposed algori...
详细信息
The paper presents a low cost skip detection algorithm and its architecture by detecting 4x4-zero block numbers in a macroblock (MB). With this simple detection mechanism and an adaptive threshold; the proposed algorithm can correctly pre-skip 82.62% of total MB encoding, and save up to 82.8% encoding time and corresponding computing power whereas maintains similar video quality because of the high accuracy prediction. Moreover, the design only occupies 0.63K gate counts so that it can be easily combined with the H.264 encoder implementation without overhead for low power and low complexity.
In this paper a modified decision tree algorithm for anomaly detection is presented. During the tree building process, densities for the outlier class are used directly in the split point determination algorithm. No a...
详细信息
ISBN:
(纸本)9781424421749
In this paper a modified decision tree algorithm for anomaly detection is presented. During the tree building process, densities for the outlier class are used directly in the split point determination algorithm. No artificial counter-examples have to be sampled from the unknown class, which yields to more precise decision boundaries and a deterministic classification result. Furthermore, the prior of the outlier class can be used to adjust the sensitivity of the anomaly detector. The proposed method combines the advantages of classification trees with the benefit of a more accurate representation of the outliers. For evaluation, we compare our approach with other state-of-the-art anomaly detection algorithms on four standard data sets including the KDD-Cup 99. The results show that the proposed method performs as well as more complex approaches and is even superior on three out of four data sets.
In this paper we propose a new architecture for detection of hybrid layered space-time block codes (LSTBC). The proposed receiver is based on a successive interference cancellation (SIC) scheme and the QR decompositio...
详细信息
In this paper we propose a new architecture for detection of hybrid layered space-time block codes (LSTBC). The proposed receiver is based on a successive interference cancellation (SIC) scheme and the QR decomposition, which leads to a suitable hardware implementation. It was designed for zero-forcing (ZF) criterion; reduced complexity is achieved by means of an adequate rearrangement of the channel matrix elements. Simulation results show that this new architecture outperforms similar, recently reported detectors in terms of bit error rate under spatially correlated channels.
This paper not only proposes the detection algorithm of traditional ring signal tones, but also researches the Color Ring Back Tone which is well popular recent years. Different testing methods are given according to ...
详细信息
This paper not only proposes the detection algorithm of traditional ring signal tones, but also researches the Color Ring Back Tone which is well popular recent years. Different testing methods are given according to the difference of music color ring tone and voice prompt color ring tone. Tested on RMTS (Real-world Multi-channel Telephone Speech) database, experiments show that the detection rates of ring tone, music color ring tone, voice prompt color ring tone and virtual speech reach 100%, 98.5%, 98% and 97.5% respectively, 98.5% totally.
Based on the characteristics study of binary decision diagrams (BDDs) of logic functions, a method using the characteristic triangle (CT) to detect logic functions suitable for XOR logic implementation is presented an...
详细信息
Based on the characteristics study of binary decision diagrams (BDDs) of logic functions, a method using the characteristic triangle (CT) to detect logic functions suitable for XOR logic implementation is presented and a corresponding algorithm is developed. The proposed algorithm is implemented in C and tested on MCNC benchmarks. The experimental results show that the proposed algorithm is efficient compared with published results.
In this paper we present an innovative approach to improve the performance of beat detection in the presence of a significant level of noise on some ECG leads in a setting of multi-lead acquisition. The idea is to run...
详细信息
In this paper we present an innovative approach to improve the performance of beat detection in the presence of a significant level of noise on some ECG leads in a setting of multi-lead acquisition. The idea is to run simultaneously two detectors (ldquoprincipalrdquo and ldquosparerdquo) and to dynamically switch between them according to the level of noise on the leads used by the detectors. First, default independent leads are assigned to the detectors; then, a bootstrap phase is necessary during which average QRS amplitudes are established for all leads. Based on the independence of the leads and the QRS amplitudes best leads are automatically computed and assigned to the detectors. After the bootstrap phase, the algorithm calculates the noise level of the leads used by both detectors every 500 ms and switches to the spare detector if the noise in the principal detector leads exceeds a threshold (T). The performance of the algorithm was tested on a database created for the purpose; the database contains 101 1-hour 12-lead Holter records selected from 24-hour recordings with a commercial analysis system on the basis of the amount of artifact. The performance of the algorithm was tested for various values of T and noise observation windows. A QRS sensitivity of 99.7 %, a QRS positive predictive value of 99.3 %, a ventricular sensitivity of 91.0 % and a ventricular positive predictive value of 55.3 % have been achieved.
暂无评论