115 Hits in 5.5 sec

Exploiting Human Social Cognition for the Detection of Fake and Fraudulent Faces via Memory Networks [article]

Tharindu Fernando, Clinton Fookes, Simon Denman, Sridha Sridharan
2019 arXiv   pre-print
We propose a Hierarchical Memory Network (HMN) architecture, which is able to successfully detect faked faces by utilising knowledge stored in neural memories as well as visual cues to reason about the  ...  and exploit the long-term dependencies stored in memory.  ...  Memory (M t-1 ) Input Controller Output Controller Update Controller Input f t q t z t r t This motivates the need for external memory components.  ... 
arXiv:1911.07844v1 fatcat:h4hhto4byjgbbnaks77slgedca

Guidelines for Artificial Intelligence Containment [article]

James Babcock, Janos Kramar, Roman V. Yampolskiy
2017 arXiv   pre-print
With almost daily improvements in capabilities of artificial intelligence it is more important than ever to develop safety software for use by the AI research community.  ...  Such safety container software will make it possible to study and analyze intelligent artificial agent while maintaining certain level of safety against information leakage, social engineering attacks  ...  image recognition, largely via advances in deep learning [18] .  ... 
arXiv:1707.08476v1 fatcat:amjutmynzvgvnf3oixageslfxu

SoK: On the Security Challenges and Risks of Multi-Tenant FPGAs in the Cloud [article]

Shaza Zeitouni, Ghada Dessouky, Ahmad-Reza Sadeghi
2020 arXiv   pre-print
We further survey and classify existing academic works that demonstrate a new class of remotely exploitable physical attacks on multi-tenant FPGA devices, where these attacks are launched remotely by malicious  ...  clients sharing physical resources with victim users.  ...  Remotely-Exploitable Physical Attacks.  ... 
arXiv:2009.13914v2 fatcat:mbdpjfuoljderjhopoppxkxkoe

India, The Fourth Industrial Revolution and Government Policy

S. Patanjali, D. Subramaniam
2019 Arthshastra Indian Journal of Economics & Research  
Artificial intelligence and machine learning capabilities are growing at an unprecedented rate.  ...  We focus instead on what sorts of attacks we are likely to see soon if adequate defenses are not developed.  ...  Another difficulty with this control point is that attackers can learn how to evade system-level defenses.  ... 
doi:10.17010/aijer/2019/v8i2/145224 fatcat:srchlppo6za4hd4klygunfpwuq

Unsupervised Machine Learning for Networking: Techniques, Applications and Research Challenges [article]

Muhammad Usama, Junaid Qadir, Aunn Raza, Hunain Arif, Kok-Lim Alvin Yau, Yehia Elkhatib, Amir Hussain, Ala Al-Fuqaha
2017 arXiv   pre-print
control (e.g., for developing autonomous self-driving cars).  ...  machine learning.  ...  [282] proposed a supervised deep learning based routing scheme for heterogeneous network traffic control.  ... 
arXiv:1709.06599v1 fatcat:llcg6gxgpjahha6bkhsitglrsm

Unsupervised Machine Learning for Networking: Techniques, Applications and Research Challenges

Muhammad Usama, Junaid Qadir, Aunn Raza, Hunain Arif, Kok-lim Alvin Yau, Yehia Elkhatib, Amir Hussain, Ala Al-Fuqaha
2019 IEEE Access  
INDEX TERMS Machine learning, deep learning, unsupervised learning, computer networks.  ...  and optimal control (e.g., for developing autonomous self-driving cars).  ...  Reference [278] proposed a supervised deep learning based routing scheme for heterogeneous network traffic control.  ... 
doi:10.1109/access.2019.2916648 fatcat:xutxh3neynh4bgcsmugxsclkna

Anomaly Detection in Blockchain Networks: A Comprehensive Survey [article]

Muneeb Ul Hassan, Mubashir Husain Rehmani, Jinjun Chen
2021 arXiv   pre-print
These anomaly detection models autonomously detect and predict anomaly in the network in order to protect network from unexpected attacks.  ...  Over the past decade, blockchain technology has attracted a huge attention from both industry and academia because it can be integrated with a large number of everyday applications working over features  ...  memory.  ... 
arXiv:2112.06089v1 fatcat:o6z2f7gunvd3fnxhfwmb57qwhm

Verification for Machine Learning, Autonomy, and Neural Networks Survey [article]

Weiming Xiang and Patrick Musau and Ayana A. Wild and Diego Manzanas Lopez and Nathaniel Hamilton and Xiaodong Yang and Joel Rosenfeld and Taylor T. Johnson
2018 arXiv   pre-print
Autonomy in CPS is enabling by recent advances in artificial intelligence (AI) and machine learning (ML) through approaches such as deep neural networks (DNNs), embedded in so-called learning enabled components  ...  (LECs) that accomplish tasks from classification to control.  ...  collected knowledge in memory.  ... 
arXiv:1810.01989v1 fatcat:a5ax66lsxbho3fuxuh55ypnm6m

Security issues in cloud environments: a survey

Diogo A. B. Fernandes, Liliana F. B. Soares, João V. Gomes, Mário M. Freire, Pedro R. M. Inácio
2013 International Journal of Information Security  
It addresses several key topics, namely vulnerabilities, threats and attacks, proposing a taxonomy for their classification.  ...  The possibility of paying-as-you-go mixed with an on-demand elastic operation is changing the enterprise computing model, shifting on-premises infrastructures to offpremises data centers, accessed over  ...  Since the appearance of Web 2.0, a new class of threats emerged along with the people learning how to exploit them.  ... 
doi:10.1007/s10207-013-0208-7 fatcat:55o67epb6zfspchxuzvuduzr4a

Conference Guide [Front matter]

2020 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV)  
For this, the use of Deep Recurrent Q-Network (DRQN) is explored, a method combining state-of-the art Deep Q-Network (DQN) with a long term short term memory (LSTM) layer helping the agent gain a memory  ...  The FDI attacks are randomly injected into communication channels or sensors with certain probabilities, which undermines the accuracy of the transmission data and the accuracy of measurement data respectively  ...  When the communication network meets the necessary connectivity, we develop a variation of so-called ratio consensus algorithm that deal with malicious attacks.  ... 
doi:10.1109/icarcv50220.2020.9305477 fatcat:4h7gpoj7ljgsrlkjoyw3qcfzxi

Overview - Fog Computing and Internet-of-Things (IOT)

C. S. R. Prabhu
2017 EAI Endorsed Transactions on Cloud Systems  
Analytics solutions such as regression techniques from Machine Learning and Deep Learning can harness the temporal dimension of IOT data.  ...  Side Channel Attack: An attacker could attempt to compromise the cloud through placing a malicious virtual machine in close proximity to target the cloud server and then exploiting a side channel attack  ... 
doi:10.4108/eai.20-12-2017.154378 fatcat:k3qhfv6ppje4lhmplwqbvcvkqi

Self-Supervised Representation Learning: Introduction, Advances and Challenges [article]

Linus Ericsson, Henry Gouk, Chen Change Loy, Timothy M. Hospedales
2021 arXiv   pre-print
of the main barriers to practical deployment of deep learning today.  ...  Self-supervised representation learning methods aim to provide powerful deep feature learning without the requirement of large annotated datasets, thus alleviating the annotation bottleneck that is one  ...  train deep representations.  ... 
arXiv:2110.09327v1 fatcat:qoprtdh4rzg6lcylgn5rafubpe

OSTP-AWhite House Office of Science and Technology Policy Request for Information on the Future of Artificial Intelligence

Kris Kitchen
2017 Figshare  
induction, security in learning source provenance, user modeling, and values modeling.  ...  ensuring that advances in AI lead to public good, and that additional federal and philanthropic research funding in these areas can potentially have very high positive impact, and should increase in pace with  ...  Similarly, Deep Mind developed a program that learned video games by practicing them.  ... 
doi:10.6084/m9.figshare.4640305 fatcat:26aelpe4ejbupao5yh2mys2pzq

Using complexity to protect elections

Piotr Faliszewski, Edith Hemaspaandra, Lane A. Hemaspaandra
2010 Communications of the ACM  
attacks.  ...  In Table 1 , I (immunity) means one can never change the outcome with that type of control attack-a dream case; R (resistance) means it is NP-hard to determine whether a given instance can be successfully  ...  So, the last thing you probably have on your mind is whether or not you are properly insured.  ... 
doi:10.1145/1839676.1839696 fatcat:hbqpm5boabe3jcpa4jcs7czf6y

On the Opportunities and Risks of Foundation Models [article]

Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, Erik Brynjolfsson, Shyamal Buch (+102 others)
2021 arXiv   pre-print
Though foundation models are based on standard deep learning and transfer learning, their scale results in new emergent capabilities,and their effectiveness across so many tasks incentivizes homogenization  ...  To tackle these questions, we believe much of the critical research on foundation models will require deep interdisciplinary collaboration commensurate with their fundamentally sociotechnical nature.  ...  Deep learning was fueled by larger datasets, more computation (notably, the availability of GPUs), and greater audacity.  ... 
arXiv:2108.07258v2 fatcat:yktkv4diyrgzzfzqlpvaiabc2m
« Previous Showing results 1 — 15 out of 115 results