scholar.google.com › citations
We propose a method to revise the neural network to construct the quaternion-valued neural network (QNN), in order to prevent intermediate-layer features from leaking input information. The QNN uses quaternion-valued features, where each element is a quaternion.
Mar 18, 2020
We propose a method to revise the neural network to construct the quaternion-valued neural network. (QNN), in order to prevent intermediate-layer fea-.
Sep 11, 2024 · The QNN hides input information into a random phase of quaternion-valued features. Even if attackers have obtained network parameters and ...
In order to prevent leaking input information from intermediate-layer features, this paper proposes a method to revise the traditional neural network into the ...
Sep 23, 2023 · This paper proposes an innovative solution called the quaternion based neural network, Advanced Data Security Architecture in Healthcare Environment (ADSAH),
Missing: Features | Show results with:Features
Oct 13, 2024 · We introduce a novel cancelable face recognition methodology by using quaternion transform based convolutional network.
We propose a method to revise the neural network to construct the quaternion-valued neural network (QNN), in order to prevent intermediate-layer features from ...
Oct 22, 2024 · To fully preserve features of color images, we introduce quaternion theory to encode each color image and propose an improved vector homomorphic ...
[PDF] Rotation-Equivariant Neural Networks for Privacy Protection ...
www.semanticscholar.org › paper
Deep Quaternion Features for Privacy Protection · Privacy-Preserving Face Recognition in the Frequency Domain · Adaptive Hybrid Masking Strategy for Privacy- ...
Deep Quaternion Features for Privacy Protection. | D2R Server ...
dblp.l3s.de › corr › abs-2003-08365
Deep Quaternion Features for Privacy Protection. Resource URI: https://dblp.l3s.de/d2r/resource/publications/journals/corr/abs-2003-08365. Home | Example ...