A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Filters
A Smooth Representation of Belief over SO(3) for Deep Rotation Learning with Uncertainty
[article]
2020
arXiv
pre-print
Preserved Fulltext
Other Versions
In this work, we present a novel symmetric matrix representation of the 3D rotation group, SO(3), with two important properties that make it particularly suitable for learned models: (1) it satisfies a ...
smoothness property that improves convergence and generalization when regressing large rotation targets, and (2) it encodes a symmetric Bingham belief over the space of unit quaternions, permitting the ...
Supervised Learning: Rotation Loss Functions For supervised learning over rotations, there are a number of possible choices for loss functions that are defined over SO (3) . ...
arXiv:2006.01031v2
fatcat:dmjy7ivojjgobcuv5ajxrbnow4
3-D Structural geological models: Concepts, methods, and uncertainties
[chapter]
2018
Advances in Geophysics
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We introduce a formulation for geological model representation and interpolation and uncertainty analysis methods with the aim to clarify similarities and differences in the diverse set of approaches that ...
We believe that some reasons for this omission are (a) an incomplete picture of available geological modeling methods, and (b) the problem of the perceived static picture of an inflexible geological representation ...
For these cases, a suitable summary measure for a representation of uncertainties in the full 3-D space is required. ...
doi:10.1016/bs.agph.2018.09.001
fatcat:drmjlawqy5c5rhotprbh7hlfv4
Temporal Difference Variational Auto-Encoder
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Motivated by the absence of a model satisfying all these requirements, we propose TD-VAE, a generative sequence model that learns representations containing explicit beliefs about states several steps ...
To act and plan in complex environments, we posit that agents should have a mental simulator of the world with three characteristics: (a) it should build an abstract state representing the condition of ...
We are therefore left with a forward encoder which ideally computes the belief state, a backwards encoder which -with the forward encoder -compute posteriors over states, and a state-to-state forward model ...
arXiv:1806.03107v3
fatcat:zokst3l3pfbjti5t6j7ttqrncu
Being Bayesian about Categorical Probability
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
As a Bayesian alternative to the softmax, we consider a random variable of a categorical probability over class labels. ...
Neural networks utilize the softmax as a building block in classification tasks, which contains an overconfidence problem and lacks an uncertainty representation ability. ...
Acknowledgements We would like to thank Dong-Hyun Lee and anonymous reviewers for the discussions and suggestions. ...
arXiv:2002.07965v2
fatcat:tzn3sspv6rheljl7jd44qzd32u
Deep Learning Point Cloud Odometry
2021
U Porto Journal of Engineering
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
of the benefits of leveraging learning-based encoding representations of real-world data is provided. ...
In addition, a broad perspective of emergent Deep Learning robust techniques to track motion and estimate scene structure for real-world applications is the focus of a deeper analysis and comprehensive ...
Acknowledgements The authors wish to acknowledge the support of the Portuguese funding agency, FCT -Fundação para a Ciência e a Tecnologia, through PhD research grant 2020.05052.BD. ...
doi:10.24840/2183-6493_007.003_0006
fatcat:orrlfkowkzgsjijnl5vnlczzzm
3D ShapeNets: A Deep Representation for Volumetric Shapes
[article]
2015
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Extensive experiments show that our 3D deep representation enables significant performance improvement over the-state-of-the-arts in a variety of tasks. ...
To this end, we propose to represent a geometric 3D shape as a probability distribution of binary variables on a 3D voxel grid, using a Convolutional Deep Belief Network. ...
We thank Thomas Funkhouser, Derek Hoiem, Alexei A. Efros, Andrew Owens, Antonio Torralba, Siddhartha Chaudhuri, and Szymon Rusinkiewicz for valuable discussion. ...
arXiv:1406.5670v3
fatcat:yyoyxqbyqjdxlifvb73x2rnrvu
3D ShapeNets: A deep representation for volumetric shapes
2015
2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2016; you can also visit the original URL.
The file type is application/pdf.
Extensive experiments show that our 3D deep representation enables significant performance improvement over the-state-of-the-arts in a variety of tasks. ...
To this end, we propose to represent a geometric 3D shape as a probability distribution of binary variables on a 3D voxel grid, using a Convolutional Deep Belief Network. ...
We thank Thomas Funkhouser, Derek Hoiem, Alexei A. Efros, Andrew Owens, Antonio Torralba, Siddhartha Chaudhuri, and Szymon Rusinkiewicz for valuable discussion. ...
doi:10.1109/cvpr.2015.7298801
dblp:conf/cvpr/WuSKYZTX15
fatcat:tkadr5auafdpvglgmsooo3ysde
A&G Volume 55 Issue 3, Full Issue
2014
Astronomy & Geophysics
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
Here again, we consider a deformable lens rather than a deformable mirror for ease of representation. ...
The light curve shows no variation for most of the year of data taking, and an upward excursion lasting over one month, with a maximum increase of ~2 mag. ...
Note that tickets for these events have been selling fast, so get in touch at once if you wish to attend. ...
doi:10.1093/astrogeo/atu134
fatcat:c5eui37ap5aophjflkzshczguy
Top–Down Connections in Self-Organizing Hebbian Networks: Topographic Class Grouping
2010
IEEE Transactions on Autonomous Mental Development
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
As summarized in [32] , other results from similar algorithms on this NORB dataset include logistic regression (19.6%), SVM with a Gaussian kernel (11.6%), and deep belief networks with a greedily pretrained ...
The visual cortex is a deep hierarchical bidirectional network [1] - [3] . ...
doi:10.1109/tamd.2010.2072150
fatcat:4rckp3y445bojipl5j3mfoqmti
Learning to track environment state via predictive autoencoding
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
This work introduces a neural architecture for learning forward models of stochastic environments. ...
The network can output both expectation over future observations and samples from belief distribution. The resulting functionalities are similar to those of a Particle Filter (PF). ...
The observations can be noisy and intermittent, the outputs include expectation over future observations, samples of future observations and belief states (as a neural representation). ...
arXiv:2112.07745v1
fatcat:ehhqc4cmyfdjlgmhra3yh4v7wi
Wind Turbine Drivetrain Expert Fault Detection System: Multivariate Empirical Mode Decomposition based Multi-sensor Fusion with Bayesian Learning Classification
2020
Intelligent Automation and Soft Computing
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
(Jing, Wang, Zhao, & Wang, 2017) proposed the deep learning networks to fuse the features learned from the raw vibration signal. ...
At learning, the cardinality is set to 0 to 3 to calculate the probability. Table 5 compares the Naïve Bayes and belief net classifiers performance. ...
doi:10.32604/iasc.2020.013924
fatcat:4dzn6xzsajb43dpwm5zb2bexsi
Autonomous Exploration Under Uncertainty via Deep Reinforcement Learning on Graphs
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We propose a novel approach that uses graph neural networks (GNNs) in conjunction with deep reinforcement learning (DRL), enabling decision-making over graphs containing exploration information to predict ...
For this problem, belief space planning methods that forward-simulate robot sensing and estimation may often fail in real-time implementation, scaling poorly with increasing size of the state, belief and ...
[19] implementation of iSAM2 [20] to perform nonlinear least-squares smoothing over a factor graph. ...
arXiv:2007.12640v1
fatcat:4x2lsdzoazeq7ayauocjsjqqq4
Greedy Deep Dictionary Learning
[article]
2016
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
We compare our results with other deep learning tools like stacked autoencoder and deep belief network; and state of the art supervised dictionary learning tools like discriminative KSVD and label consistent ...
In this work we propose a new deep learning tool called deep dictionary learning. Multi-level dictionaries are learnt in a greedy fashion, one layer at a time. ...
Comparison with other Deep Learning Approaches We compared our results with a stacked autoencoder (SAE) and deep belief network (DBN). ...
arXiv:1602.00203v1
fatcat:otji3ovu75aijlcji3azh7k3wa
Vol. 2, No. 3, August 2009
2009
International Education Studies
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
Process NGT is a method that allows ideas to be generated in relation to a situation that is surrounded with uncertainty. ...
requires a rich and deep understanding, so teacher behaviors and characteristics that support teacher clarity are valued. ...
doi:10.5539/ies.v2n3p0
fatcat:3smidj2gcbd3jlzg6dt56y5e3y
Active Robotic Mapping through Deep Reinforcement Learning
[article]
2017
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
In contrast to prior work, this approach learns an exploration policy based on a user-specified prior over environment configurations and sensor model, allowing it to specialize to the specifications. ...
We propose an approach to learning agents for active robotic mapping, where the goal is to map the environment as quickly as possible. ...
First, we require a representation of the map as well as our belief over possible maps. Second, we need a way to update our belief over maps given observations. ...
arXiv:1712.10069v1
fatcat:uxkywyzw3zcdtgnsjyd5hpcf74
« Previous
Showing results 1 — 15 out of 5,702 results