Filters








8 Hits in 8.4 sec

W-Trans: A Weighted Transition Matrix Learning Algorithm for the Sensor-based Human Activity Recognition

Changhai Wang, Bo Wang, Hui Liang, Jianzhong Zhang, Wanwei Huang, Wangwei Zhang, Hongxue Shen
2020 IEEE Access  
Aiming at the problem that the current classic transition matrix learning algorithm cannot fuse weights of sample classification results, a weighted transition matrix learning algorithm is proposed in  ...  The sensor-based human activity recognition has been wildly applied in behavior tracking, health monitoring, indoor localization etc.  ...  Based on the definitions of WOHMM, the weighted transition matrix learning algorithm W-Trans is proposed.  ... 
doi:10.1109/access.2020.2984456 fatcat:ijnqnxtvmnellntyiqz5opbij4

Dense semantic SLAM

Renato F. Salas-Moreno, Andrew Davison
2015
for robust real-time camera tracking that relies on a consistent map.  ...  The created maps however were designed for the sole purpose of localising the robot (i.e. what is the position and orientation of the robot in relation to the map) and several systems demonstrated the  ...  We use the following inverse prior covariances: Σ −1 o j ,f = diag(0, w trans , 0, w rot , 0, w rot ) , (5.19) with positive weights w trans and w rot .  ... 
doi:10.25560/24524 fatcat:jdmx36lo2fd6toedwzjlcnebnq

Multi-View Signal Processing and Learning on Graphs

Xiaowen Dong
2014
For example, the works in [101, 69, 20] propose to use functional maps (namely, a given set of corresponding functions) to replace the assumption on the bijective correspondences between vertices in  ...  for properly representing the data.  ...  Then, the state at the next time stamp t + 1 is computed as: p(t + 1) = W tran T p(t), (5.3) where W tran = D 1 W is the transition probability matrix.  ... 
doi:10.5075/epfl-thesis-6213 fatcat:ph62eycbrre4vlulaoz7q6taj4

Soccer Playing Humanoid Robots

Christian Albrecht, Manuel Barbi, Karen Bieling, Torsten Böttinger, Marco Dürr, Lennart Downar, Marcel Eilers, Julian Hannich, Janine Hemmers, Fabian Rensen, Heiner Walter, Florian Ziegler (+2 others)
2016
Both single weights w(∆x) and w(∆α) are combined by the weighted mean w(∆x, ∆α) = w trans · w(∆x) + w rot · w(∆α) w trans + w rot . (12.3) The parameters σ, w min , w trans and w rot have to be adjusted  ...  Thereafter the weights are calculated for each sample by evaluating the sensor information (sensor update).  ...  It is possible to create a CABSL-Editor as an Eclipse-Plugin which offers Syntax-Highlighting, an Outline and Auto-Completion for easy development of new CABSL-Options.  ... 
doi:10.17877/de290r-17674 fatcat:u4w4zrlpdfen5hxfqxkfmfcjdu

One-class classification in the presence of point, collective, and contextual anomalies [article]

Nico Görnitz, Technische Universität Berlin, Technische Universität Berlin, Klaus-Robert Müller
2019
One-class classification is a machine learning concept that is especially suited for the anomaly detection problem.  ...  \kmeans as a special case, and (iii) a method for learning with latent class dependencies when an additional str [...]  ...  To better understand the above feature map, observe that the weight vector w, which is w = (w em , w trans ), decomposes into a transition vector w trans = (w trans i,j ) i,j∈Y and an emission vector w  ... 
doi:10.14279/depositonce-7897 fatcat:bgrlfa5pzfacxhxonrjlqjdqza

Longitudinal Flying Qualities Criteria for Single-Pilot Instrument Flight Operations

Aharon Bar-Gill, Robert F. Stengel
1983 SAE Technical Paper Series   unpublished
the ~l([SI_F]-l) algorithm and using matrix partitioning and block diagonalization (Ref. 22) , the state transition matrix is obtained (.  ...  The state vector x used here contains 2-1 three components each of translational rate (u,v,w), trans- lational position (xI' YI' zI)' angular rate (p, q, r) and angular attitude (cI>, 8 , 1/1).  ...  Most of the configurations were flown E-I twice. In some cases, when objective data recording problems occured due to· checklist errors, a configuration was flown a third time.  ... 
doi:10.4271/830761 fatcat:7cfe4jbh7zbbjo6eyytkcgwfqa

Noninvasive imaging

2001 Journal of the American College of Cardiology  
Conclusion: CMRA using a free breathing, T2 weighted, 3DGRE navigator g&d technique is highly sensitive for the detection of significant CAD.  ...  In a subgroup of the pts with sinus rhythm (NSR) the LAA-EF was significantly less in pts with clot (31.2 + 13.2 Vs 41.2 + 11.5% p50% diameter stenosis.  ...  Weights based on activity, motion and consistency of the endo-and epicardial surface estimates III space and time are assigned.  ... 
doi:10.1016/s0735-1097(01)80006-1 fatcat:ghklt3w3zze3le5gwnpbbfvrrm

Roller Compaction of Pharmaceutical Ingredients : on the Undrestanding of the Compaction and the Use of Knowledge based Applications in the Formulation of Tablets

Andrés-Miguel Miguélez-Morán
2009
Fricker for accepting me as a doctorate student at the institute, for allowing me to perform the major part of the experimental of this work, and for agreeing to be my supervisor and my referee.  ...  I also thank him the support and the effort that he has made at arranging the funding of my stay at the department during the last year and a half.  ...  W tot is the total work exerted on the powder bed and W trans is the amount of work transferred to the system.  ... 
doi:10.11588/heidok.00010111 fatcat:4gchgcaq2jcxfol64mwyikghdq