Fully Hyperbolic Convolutional Neural Networks for Computer Vision
- URL: http://arxiv.org/abs/2303.15919v3
- Date: Wed, 7 Feb 2024 13:46:35 GMT
- Title: Fully Hyperbolic Convolutional Neural Networks for Computer Vision
- Authors: Ahmad Bdeir and Kristian Schwethelm and Niels Landwehr
- Abstract summary: We present HCNN, a fully hyperbolic convolutional neural network (CNN) designed for computer vision tasks.
Based on the Lorentz model, we propose novel formulations of the convolutional layer, batch normalization, and multinomial logistic regression.
Experiments on standard vision tasks demonstrate the promising performance of our HCNN framework in both hybrid and fully hyperbolic settings.
- Score: 3.3964154468907486
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Real-world visual data exhibit intrinsic hierarchical structures that can be
represented effectively in hyperbolic spaces. Hyperbolic neural networks (HNNs)
are a promising approach for learning feature representations in such spaces.
However, current HNNs in computer vision rely on Euclidean backbones and only
project features to the hyperbolic space in the task heads, limiting their
ability to fully leverage the benefits of hyperbolic geometry. To address this,
we present HCNN, a fully hyperbolic convolutional neural network (CNN) designed
for computer vision tasks. Based on the Lorentz model, we generalize
fundamental components of CNNs and propose novel formulations of the
convolutional layer, batch normalization, and multinomial logistic regression.
{Experiments on standard vision tasks demonstrate the promising performance of
our HCNN framework in both hybrid and fully hyperbolic settings.} Overall, we
believe our contributions provide a foundation for developing more powerful
HNNs that can better represent complex structures found in image data. Our code
is publicly available at https://github.com/kschwethelm/HyperbolicCV.
Related papers
- Fully Spiking Actor Network with Intra-layer Connections for
Reinforcement Learning [51.386945803485084]
We focus on the task where the agent needs to learn multi-dimensional deterministic policies to control.
Most existing spike-based RL methods take the firing rate as the output of SNNs, and convert it to represent continuous action space (i.e., the deterministic policy) through a fully-connected layer.
To develop a fully spiking actor network without any floating-point matrix operations, we draw inspiration from the non-spiking interneurons found in insects.
arXiv Detail & Related papers (2024-01-09T07:31:34Z) - Heterogeneous Graph Convolutional Neural Network via Hodge-Laplacian for
Brain Functional Data [4.80657982213439]
This study proposes a novel heterogeneous graph convolutional neural network (HGCNN) to handle complex brain fMRI data.
We introduce a generic formulation of spectral filters on heterogeneous graphs by introducing the $k-th$lacian (HL) operator.
We design HL-node, HL-edge, and HL-HGCNN neural networks to learn signal representation at a graph node, edge levels, and both, respectively.
arXiv Detail & Related papers (2023-02-18T12:58:50Z) - Hybrid SNN-ANN: Energy-Efficient Classification and Object Detection for
Event-Based Vision [64.71260357476602]
Event-based vision sensors encode local pixel-wise brightness changes in streams of events rather than image frames.
Recent progress in object recognition from event-based sensors has come from conversions of deep neural networks.
We propose a hybrid architecture for end-to-end training of deep neural networks for event-based pattern recognition and object detection.
arXiv Detail & Related papers (2021-12-06T23:45:58Z) - ACE-HGNN: Adaptive Curvature Exploration Hyperbolic Graph Neural Network [72.16255675586089]
We propose an Adaptive Curvature Exploration Hyperbolic Graph NeuralNetwork named ACE-HGNN to adaptively learn the optimal curvature according to the input graph and downstream tasks.
Experiments on multiple real-world graph datasets demonstrate a significant and consistent performance improvement in model quality with competitive performance and good generalization ability.
arXiv Detail & Related papers (2021-10-15T07:18:57Z) - Improvising the Learning of Neural Networks on Hyperspherical Manifold [0.0]
The impact of convolution neural networks (CNNs) in the supervised settings provided tremendous increment in performance.
The representation learned from CNN's operated on hyperspherical manifold led to insightful outcomes in face recognition.
A broad range of activation functions is developed with hypersphere intuition which performs superior to softmax in euclidean space.
arXiv Detail & Related papers (2021-09-29T22:39:07Z) - Free Hyperbolic Neural Networks with Limited Radii [32.42488915688723]
Hyperbolic Neural Networks (HNNs) that operate directly in hyperbolic space have been proposed recently to further exploit the potential of hyperbolic representations.
While HNNs have achieved better performance than Euclidean neural networks (ENNs) on datasets with implicit hierarchical structure, they still perform poorly on standard classification benchmarks such as CIFAR and ImageNet.
In this paper, we first conduct an empirical study showing that the inferior performance of HNNs on standard recognition datasets can be attributed to the notorious vanishing gradient problem.
Our analysis leads to a simple yet effective solution called Feature Clipping, which regularizes the hyperbolic embedding whenever its
arXiv Detail & Related papers (2021-07-23T22:10:16Z) - Fully Hyperbolic Neural Networks [63.22521652077353]
We propose a fully hyperbolic framework to build hyperbolic networks based on the Lorentz model.
We show that our method has better performance for building both shallow and deep networks.
arXiv Detail & Related papers (2021-05-31T03:36:49Z) - Hyper-Convolution Networks for Biomedical Image Segmentation [22.902923145462008]
The size of the convolution kernels determines both the expressiveness of convolutional neural networks (CNN) and the number of learnable parameters.
We propose a powerful novel building block, the hyper-convolution, which implicitly represents the convolution kernel as a function of kernel coordinates.
We demonstrate that replacing regular convolutions with hyper-convolutions leads to more efficient architectures that achieve improved accuracy.
arXiv Detail & Related papers (2021-05-21T20:31:08Z) - Spatial Dependency Networks: Neural Layers for Improved Generative Image
Modeling [79.15521784128102]
We introduce a novel neural network for building image generators (decoders) and apply it to variational autoencoders (VAEs)
In our spatial dependency networks (SDNs), feature maps at each level of a deep neural net are computed in a spatially coherent way.
We show that augmenting the decoder of a hierarchical VAE by spatial dependency layers considerably improves density estimation.
arXiv Detail & Related papers (2021-03-16T07:01:08Z) - Hyperbolic Generative Adversarial Network [0.0]
We propose that it is possible to take advantage of the hierarchical characteristic present in the images by using hyperbolic neural networks in a GAN architecture.
In this study, different configurations using fully connected hyperbolic layers in the GAN, CGAN, and WGAN are tested, in what we call the HGAN, HCGAN, and HWGAN, respectively.
arXiv Detail & Related papers (2021-02-10T16:55:27Z) - Binary Graph Neural Networks [69.51765073772226]
Graph Neural Networks (GNNs) have emerged as a powerful and flexible framework for representation learning on irregular data.
In this paper, we present and evaluate different strategies for the binarization of graph neural networks.
We show that through careful design of the models, and control of the training process, binary graph neural networks can be trained at only a moderate cost in accuracy on challenging benchmarks.
arXiv Detail & Related papers (2020-12-31T18:48:58Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.