Results 81 to 90 of about 29,955 (264)
Interpreting Adversarial Examples with Attributes
CoRR, 2019 Deep computer vision systems being vulnerable to imperceptible and carefully crafted noise have raised questions regarding the robustness of their decisions. We take a step back and approach this problem from an orthogonal direction. We propose to enable black-box neural networks to justify their reasoning both for clean and for adversarial examples by
Sadaf Gulshad +3 more
openaire +3 more sources
Advanced Science, EarlyView.Deep‐learning‐based signal enhancement is an effective way to recover high‐resolution details from a low‐resolution chromatin contact map. However, due to computational challenges, existing methods commonly divide up the contact map into small patches and create artificial discontinuities at patch boundaries.
Qinyao Li +6 more
wiley +1 more source
On the (Statistical) Detection of Adversarial Examples
CoRR, 2017 Machine Learning (ML) models are applied in a variety of tasks such as network intrusion detection or Malware classification. Yet, these models are vulnerable to a class of malicious inputs known as adversarial examples. These are slightly perturbed inputs that are classified incorrectly by the ML model.
Kathrin Grosse +4 more
openaire +2 more sources
Sustainable Materials Design With Multi‐Modal Artificial Intelligence
Advanced Science, EarlyView.Critical mineral scarcity, high embodied carbon, and persistent pollution from materials processing intensify the need for sustainable materials design. This review frames the problem as multi‐objective optimization under heterogeneous, high‐dimensional evidence and highlights multi‐modal AI as an enabling pathway.
Tianyi Xu +8 more
wiley +1 more source
Advanced Science, EarlyView.A latent diffusion‐based framework is proposed for designing functionally graded metamaterials with perfect connectivity. By integrating vector‐quantized latent representations with mechanistic guidance, the framework enables accurate inverse design toward target elastic properties.
Jongbin Yu, Dosung Lee, Namjung Kim
wiley +1 more source
Adversarial robustness analysis of LiDAR-included models in autonomous driving
High-Confidence ComputingIn autonomous driving systems, perception is pivotal, relying chiefly on sensors like LiDAR and cameras for environmental awareness. LiDAR, celebrated for its detailed depth perception, is being increasingly integrated into autonomous vehicles.
Bo Yang +4 more
doaj +1 more source
Advanced Science, EarlyView.Technical limitations often let dominant signals overshadow rare cell types and fine‐grained heterogeneity in spatial transcriptomics. SemanticST, a graph neural network using multi‐semantic graph fusion and a novel min‐cut loss, recovers these subtle patterns.
Roxana Zahedi +7 more
wiley +1 more source
Provably Robust Adversarial Examples
, 2020 International Conference on Learning Representations (ICLR 2022)
Dimitar Iliev Dimitrov +3 more
openaire +4 more sources
A New Kind of Adversarial Example
CoRR, 2022 Almost all adversarial attacks are formulated to add an imperceptible perturbation to an image in order to fool a model. Here, we consider the opposite which is adversarial examples that can fool a human but not a model. A large enough and perceptible perturbation is added to an image such that a model maintains its original decision, whereas a human ...
openaire +2 more sources
Neuromorphic Devices and Computing for Sensing, Memory, and Control
Advanced Science, EarlyView.This review introduces neuromorphic devices made from diverse materials. These devices mimic neuronal functions and architectures and, when integrated with artificial or biological computing, can form closed loops with neurons for pressure, optical, acoustic, and biochemical sensing and modulation.
Zhengguang Zhu +2 more
wiley +1 more source