12,555 Hits in 7.7 sec

Opening the Black Box: Hierarchical Sampling Optimization for Estimating Human Hand Pose

Danhang Tang, Jonathan Taylor, Pushmeet Kohli, Cem Keskin, Tae-Kyun Kim, Jamie Shotton
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
We address the problem of hand pose estimation, formulated as an inverse problem. Typical approaches optimize an energy function over pose parameters using a 'black box' image generation procedure.  ...  Our new framework, hierarchical sampling optimization, consists of a sequence of predictors organized into a kinematic hierarchy.  ...  Introduction Estimating the articulated pose of the human hand presents unique challenges for computer vision algorithms.  ... 
doi:10.1109/iccv.2015.380 dblp:conf/iccv/TangTKKKS15 fatcat:pzzifwaty5f7npqvyvded37h5q

Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR [article]

Sandra Wachter, Brent Mittelstadt, Chris Russell
2018 arXiv   pre-print
Implementing a right to explanation that opens the black box of algorithmic decision-making faces major legal and technical barriers.  ...  Even though such interpretability is of great importance and should be pursued, explanations can, in principle, be offered without opening the black box.  ...  We compared these aims with the provisions of the GDPR and evaluated if they rely upon opening the black box.  ... 
arXiv:1711.00399v3 fatcat:xjhwo7dvpvcmjhswo3az3wjbwq

Opening the Black Box: Strategies for Increased User Involvement in Existing Algorithm Implementations

Thomas Muhlbacher, Harald Piringer, Samuel Gratzl, Michael Sedlmair, Marc Streit
2014 IEEE Transactions on Visualization and Computer Graphics  
In many cases, however, the algorithms are used as black boxes that run to completion in isolation which contradicts the needs of interactive data exploration.  ...  for the design of algorithmic APIs.  ...  Thanks go to Clemens Arbesser for valuable discussions, and to Johanna Schlereth for help with the figures.  ... 
doi:10.1109/tvcg.2014.2346578 pmid:26356878 fatcat:pvhmwutstjdtveterbmijk7c6q

Black-box model risk in finance [article]

Samuel N. Cohen and Derek Snow and Lukasz Szpruch
2021 arXiv   pre-print
The difficulty of understanding the inner workings of these systems, combined with their wide applicability, has the potential to lead to significant new risks for users; these risks need to be understood  ...  Our aim will be to highlight the various sources of risk that the introduction of machine learning emphasises or de-emphasises, and the possible risk mitigation and management strategies that are available  ...  However, these results suggest that more complex black-box models may fail to outperform simpler ones.  ... 
arXiv:2102.04757v1 fatcat:dwtvo4lolfe6hkb7bwswunz7lu

Analysis of Explainers of Black Box Deep Neural Networks for Computer Vision: A Survey [article]

Vanessa Buhrmester, David Münch, Michael Arens
2019 arXiv   pre-print
black boxes.  ...  As a black box model due to their multilayer nonlinear structure, Deep Neural Networks are often criticized to be non-transparent and their predictions not traceable by humans.  ...  Conclusion In this paper we give reasons why it is necessary to comprehend black box Deep Neural Networks: Adverserial attacks that are not understandable by humans pose a serious threat to the robustness  ... 
arXiv:1911.12116v1 fatcat:qgeg6rz6qzgrfikhsgah77yz2a

Digital Public History Inside and Outside the Box

Enrica Salvatori
2020 magazén  
conferences, the place of history in the large digital humanities environment and finally the definition itself of digital history.  ...  These experiments, studies and conferences have built a rich basis for the relationship between the disciplines, but the difficult connection needs to be further explored.  ...  These methods, when they are first operationalised in the nineties, function more or less as black boxes.  ... 
doi:10.30687/mag/2724-3923/2020/02/003 fatcat:dmabtb22vrfahnkayqczgorwj4

Opening the machine learning black box with Layer-wise Relevance Propagation [article]

Sebastian Lapuschkin, Technische Universität Berlin, Technische Universität Berlin, Klaus-Robert Müller
However, due to the nested complex and non-linear structure of many machine learning models, this comes with the disadvantage of them acting as a black box, providing little or no information about the  ...  This black box character hampers acceptance and application of non-linear methods in many application domains, where understanding individual model predictions and thus trust in the model's decisions are  ...  Acknowledgements Experimental Setup We demonstrate the results on a classifier for the MIT Places data set (Zhou et al., 2014b) provided by the authors of this data set and the Caffe reference model  ... 
doi:10.14279/depositonce-7942 fatcat:2ca4ccz5bzb2lojrqrpsxhmzve

Unbox the black-box for the medical explainable AI via multi-modal and multi-centre data fusion: A mini-review, two showcases and beyond

Guang Yang, Qinghao Ye, Jun Xia
2021 Information Fusion  
Consequently, our confidence in AI systems can be hindered by the lack of explainability in these black-box models.  ...  Explainable Artificial Intelligence (XAI) is an emerging research topic of machine learning aimed at unboxing how AI systems' black-box choices are made.  ...  This black-box design poses an obstacle for the validation of the developed AI algorithms.  ... 
doi:10.1016/j.inffus.2021.07.016 pmid:34980946 pmcid:PMC8459787 fatcat:3rmzvn72dbgglcddgolce2xsfe

Opening the black box of heterogeneous value creation: cognitive microfoundations of novel and valuable solution generation

Leif Willard Lundmark
for future research investigating models of cognition within the field of strategy.  ...  Microfoundational research investigating the cognitive processes of value creation effectively repositions the strategist at the center of strategic management.  ...  Thanks also to Bryan Bonner for his guidance in experimental methodologies as well as the Management Department at the University of Utah for the use of the participant pool.  ... 
doi:10.26053/0h-y1k5-ef00 fatcat:ztyxwjv55rhp3hbdae5273mwq4

Prising open the black box: The production of knowledge on the mental health 'treatment gap' in Africa

S Cooper
I thus prise open the 'black box' of this knowledge, scrutinising its underlying meaning codes and capillaries of power.  ...  To sum up, I surface particular tensions and hold them up to the light, with the hope of disturbing certain intellectual reflexes and creating a space for potential alternatives.  ...  As Bowker and Star (1999, p. 161) , assert: Black boxes may be opened and closed as circumstances and structural conditions change…the box, if you will, is neither clearly closed nor completely black.  ... 
doi:10.17037/pubs.02531621 fatcat:o2utotkwjzbttlijjgaebs6na4

Whole-Body Human Pose Estimation in the Wild [article]

Sheng Jin, Lumin Xu, Jin Xu, Can Wang, Wentao Liu, Chen Qian, Wanli Ouyang, Ping Luo
2020 arXiv   pre-print
This paper investigates the task of 2D human whole-body pose estimation, which aims to localize dense landmarks on the entire human body including face, hands, body, and feet.  ...  Extensive experiments show that COCO-WholeBody not only can be used to train deep models from scratch for whole-body pose estimation but also can serve as a powerful pre-training dataset for many different  ...  COCO-WholeBody enables us to take into account the hierarchical structure of the human body and the correlations between different body parts to estimate the entire body pose.  ... 
arXiv:2007.11858v1 fatcat:rwjd2cwbovhgvd5ifz3cqw2ihy

Keys-to-Sim: Transferring Hand-Crafted Key-framed Animations to Simulated Figures using Wide Band Stochastic Trajectory Optimization

Dominik Borer, Martin Guay, Robert W. Sumner
2018 Computer Graphics and Visual Computing  
In this paper, we address the problem of transferring hand-crafted kinematic motions to a fully simulated figure, by computing open-loop controls necessary to reproduce the target motion.  ...  throughout the optimization.  ...  Motivated by these problems, several works focused on modelfree and gradient-free optimization techniques, where the simulator (in maximal coordinates) is treated as a black box [LYvdP * 10, ABdLH13, HL14  ... 
doi:10.2312/cgvc.20181205 dblp:conf/tpcg/BorerGS18 fatcat:hffujrgda5c6jcznzwvwsckhl4

Herb 2.0: Lessons Learned From Developing a Mobile Manipulator for the Home

Siddhartha S. Srinivasa, Dmitry Berenson, Maya Cakmak, Alvaro Collet, Mehmet R. Dogar, Anca D. Dragan, Ross A. Knepper, Tim Niemueller, Kyle Strabala, Mike Vande Weghe, Julius Ziegler
2012 Proceedings of the IEEE  
In this paper, we reveal some of the structure present in everyday environments that we have been able to harness for manipulation and interaction, comment on the particular challenges on working in human  ...  We have developed HERB 2.0 to perform useful tasks for and with people in human environments.  ...  An idea for merging optimization and randomized sampling is using CHOMP as the extension operator for the tree.  ... 
doi:10.1109/jproc.2012.2200561 fatcat:uq2loi6okng6lasqrxerc4qn5e

Tracking humans interacting with the environment using efficient hierarchical sampling and layered observation models

Jan Bandouch, Michael Beetz
2009 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops  
First, a sophisticated hierarchical sampling strategy for recursive Bayesian estimation that combines partitioning with annealing strategies to enable efficient search in the presence of many local maxima  ...  Our system is capable of tracking a highdimensional human model (51 DOF) without constricting the type of motion and the need for training sequences.  ...  Hierarchical Sampling Strategy Due to the non-linearity in both motion and observation models, particle filters [8] are a common choice for estimating the pdf.  ... 
doi:10.1109/iccvw.2009.5457532 dblp:conf/iccvw/BandouchB09 fatcat:fdojd33nqfdfzpppjfkvyj5ee4

Object Manipulation with an Anthropomorphic Robotic Hand via Deep Reinforcement Learning with a Synergy Space of Natural Hand Poses

Patricio Rivera, Edwin Valarezo Añazco, Tae-Seong Kim
2021 Sensors  
The DRL policy accesses the synergistic representation and derives natural hand poses through a deep regressor for object grasping and relocation tasks.  ...  A synergy space is created using a continuous normalizing flow network with point clouds of haptic areas, representing natural hand poses obtained from human grasping demonstrations.  ...  Synergy-Based Regressor The regressor estimates the corresponding joint angles for the hand pose sampled from the low-dimensional synergy space.  ... 
doi:10.3390/s21165301 pmid:34450741 pmcid:PMC8400557 fatcat:gfoj7tgm6faalkekdsohwzy7la
« Previous Showing results 1 — 15 out of 12,555 results