Certifiers make neural networks vulnerable to availability attacks

Certifiers make neural networks vulnerable to availability attacks

To achieve reliable, robust, and safe AI systems, it is vital to implement fallback strategies when AI predictions cannot be trusted. Certifiers for neural networks are a reliable way to check the robustness of these predictions. They guarantee for some predictions that a certain class of manipulati...

Volledige beschrijving

Bibliografische gegevens
Hoofdauteurs:	Lorenz, T, Kwiatkowska, M, Fritz, M
Formaat:	Conference item
Taal:	English
Gepubliceerd in:	Association for Computing Machinery 2023

Gelijkaardige items

FullCert: deterministic end-to-end certification for training and inference of neural networks
door: Lorenz, T, et al.
Gepubliceerd in: (2024)

Certified Robustness to Text Adversarial Attacks by Randomized [MASK]
door: Jiehang Zeng, et al.
Gepubliceerd in: (2023-06-01)

Attack Vulnerability of Network Controllability.
door: Zhe-Ming Lu, et al.
Gepubliceerd in: (2016-01-01)

Bayesian inference with certifiable adversarial robustness
door: Wicker, M, et al.
Gepubliceerd in: (2021)

Vulnerability analysis on noise-injection based hardware attack on deep neural networks
door: Liu, Wenye, et al.
Gepubliceerd in: (2020)

Uncertainty in vulnerability of networks under attack
door: Alireza Ermagun, et al.
Gepubliceerd in: (2023-02-01)

Towards efficient and certified recovery from poisoning attacks in federated learning
door: Jiang, Yu, et al.
Gepubliceerd in: (2025)

CNN-Cert: An Efficient Framework for Certifying Robustness of Convolutional Neural Networks
door: Boopathy, Akhilan, et al.
Gepubliceerd in: (2021)

Attacks, vulnerabilities and security requirements in smart metering networks
door: Abdullah, Muhammad Daniel Hafiz, et al.
Gepubliceerd in: (2015)

Classification of Attacks in Simulation Studies of Complex Networks Vulnerabilities
door: A. A. Tikhomirov, et al.
Gepubliceerd in: (2012-03-01)

Reachability computation for switching diffusions: finite abstractions with certifiable and tuneable precision
door: Laurenti, L, et al.
Gepubliceerd in: (2017)

Reachability computation for switching diffusions: Finite abstractions with certifiable and tuneable precision
door: Laurenti, L, et al.
Gepubliceerd in: (2017)

Backdoor attacks in neural networks
door: Low, Wen Wen
Gepubliceerd in: (2023)

Backdoor attacks in neural networks
door: Liew, Sher Yun
Gepubliceerd in: (2024)

Interdependency and vulnerability of multipartite networks under target node attacks
door: Cai, Qing, et al.
Gepubliceerd in: (2020)

Detection of vulnerability attack through wireless network on android platform
door: Muhamad Izhan Fakri, Imran
Gepubliceerd in: (2014)

Interdependency and Vulnerability of Multipartite Networks under Target Node Attacks
door: Qing Cai, et al.
Gepubliceerd in: (2019-01-01)

Automated Segmentation to Make Hidden Trigger Backdoor Attacks Robust against Deep Neural Networks
door: Saqib Ali, et al.
Gepubliceerd in: (2023-04-01)

An assessment of the availability of cavities for secondary cavity-nesting birds in certified and conventionally-logged Neotropical rainforests
door: Schaaf AA, et al.
Gepubliceerd in: (2020-08-01)

Demystifying adversarial attacks on neural networks
door: Yip, Lionell En Zhi
Gepubliceerd in: (2020)

Adversarial attack defences for neural network
door: Singh Kirath
Gepubliceerd in: (2022)

Adversarial attack defenses for neural networks
door: Puah, Yi Hao
Gepubliceerd in: (2024)

Graph theory analysis reveals an assortative pain network vulnerable to attacks
door: Chong Chen, et al.
Gepubliceerd in: (2023-12-01)

Examining indicators of complex network vulnerability across diverse attack scenarios
door: Ahmad F. Al Musawi, et al.
Gepubliceerd in: (2023-10-01)

Vulnerability of Complex Networks under Approximate Longest Chain Attack Strategies
door: Hao Yao-Hui, et al.
Gepubliceerd in: (2016-01-01)

Certified grasping
door: Aceituno-Cabezas, Bernardo, et al.
Gepubliceerd in: (2024)

Certifiable AI
door: Jobst Landgrebe
Gepubliceerd in: (2022-01-01)

Certified domination
door: Magda Dettlaff, et al.
Gepubliceerd in: (2020-01-01)

Certified Computation
door: Arkoudas, Konstantine
Gepubliceerd in: (2004)

Safety verification for deep neural networks with provable guarantees
door: Kwiatkowska, M
Gepubliceerd in: (2019)

Embedded Systems and their Vulnerabilities to Hardware Attacks
door: Konrad Szczepankiewicz, et al.
Gepubliceerd in: (2023-03-01)

Vulnerability of blockchain technologies to quantum attacks
door: Joseph J. Kearney, et al.
Gepubliceerd in: (2021-07-01)

Indoor environmental quality in WELL-certified and LEED-certified buildings
door: Michael G. Kent, et al.
Gepubliceerd in: (2024-07-01)

Code Switching Used by Certified and Non-Certified English Teachers
door: Selvianillah Azis, et al.
Gepubliceerd in: (2023-04-01)

SIL Loop: Certified or Non Certified Equipment — The Way to Go!
door: I P Parry, et al.
Gepubliceerd in: (2002-12-01)

Vulnerability of deep neural networks for detecting COVID-19 cases from chest X-ray images to universal adversarial attacks.
door: Hokuto Hirano, et al.
Gepubliceerd in: (2020-01-01)

Assessment of water availability vulnerability in the Cerrado
door: Daniel Althoff, et al.
Gepubliceerd in: (2021-10-01)

Investigating vulnerability of watermarking neural network
door: Chua, Viroy Sheng Yang
Gepubliceerd in: (2020)

National Board-certified Teachers: Can They Make a Difference in Maine Schools?
door: Sarah V Mackenzie, et al.
Gepubliceerd in: (2008-01-01)

Robustness guarantees for deep neural networks on videos
door: Kwiatkowska, M, et al.
Gepubliceerd in: (2020)