【论文笔记】Fully Hyperbolic Convolutional Neural Networks
一篇存货

2 Fully Hyperbolic Convolutional Neural Networks for Computer Vision
https://openreview.net/forum?id=ekz1hN5QNh
一些问题
- Discuss the motivation behind using the Lorentz model instead of the Poincaré ball in this paper. What benefits does it offer? 庞加莱球模型接近边界时的距离函数是无界的,不利于计算((Bdeir et al., 2024, p. 2) (pdf))| 洛伦兹模型中的距离、度量、平移等具有形式较好的解析解、经过零点的测地线到其切空间的投影均为直线、相同维数不同曲率的双曲空间共用一个零点 分母为零,x=y时 cosh(1) cosh(10)
- The paper introduces Lorentzian batch normalization. Describe how it generalizes traditional batch normalization for hyperbolic space.
一般的BN层包括减去
和线性放缩两步。而Frechet平均没有解析解,会拖慢模型训练进度。文中提出使用平方双曲距离的centroid作为Frechet平均的一个近似。然后利用零点处测地线投影为直线的性质将双曲空间的点投影至centroid处的切空间后平移至零点切空间,在此进行放缩之后再移回去((Bdeir et al., 2024, p. 5) (pdf)) - What is the Lorentzian centroid, and why is it used in batch normalization?
Lorentzian centroid以下优化问题的解析解
。通过最小化至各点平方双曲距离之和作为Frechet平均的一个近似,而后者难以在短时间内通过迭代得到足够好的逼近。 - Derive and explain the formula for the distance between a point and a hyperplane in the Lorentz model as presented in the paper. 附录写了,双曲空间中的点可以关于超平面获取其镜像。该点和镜像的测地线段等于该点至超平面距离的两倍。
- Review the implementation of the Lorentz convolutional layer in the code repository. How does it handle feature maps and kernel operations? 将每个要做卷积的kernel提出来,然后reshape好,对所有的patches做lorentz fully connected运算
- How is parallel transport used in the code, and why is it critical for operations in hyperbolic space? 用于BN层和Wrapped Normal分布(VAE里面用的,但我还没看)。 对于双曲空间中非零点的切空间,其测地线在切空间上的投影是曲线,不好控制。
- Examine how the paper implements residual connections in the Lorentz model. What challenges arise, and how are they addressed? 当stride不为1或输出维数不为预测维数时会产生形状不对齐的情况,作者采用的方法是使用1x1卷积带上stride参数进行调整
- If you were to improve the HCNN framework, what additional components or optimizations would you suggest? 考虑Attention?(类似第一篇文章)优化方面考虑在除了pytorch模块运行之外的双曲空间上的若干操作,如指数对数映射,parallel transport是否消耗太多时间,是否可以使用差一些精度的近似来提高速度
Hyperbolic Geometry
The
Poincaré ball The
Lorentz Model The
with the Lorentzian inner product
We call the first component the time component

Distance on Lorentz Model The shortest path on the model is inherited from Minkowski space. Let
When calculating distance to origin
Tangent Space on Lorentz Model The space around each point
Exponential and Logarithmic Maps on Lorentz Model Exponential and logarithmic maps between manifold
When dealing with tangent space at origin
Fully Hyperbolic CNN
Hyperbolic Feature Maps Adding Lotentz model's time component as additional channel dimension, which difines a hyperbolic feature mao as an ordered set of
Lorentz Fully-connected Layer Let
Adding other compo nents including normalization, the Lorentz fully connected layer becomes
in which
If normalization is removed, the Lorentz fully-connected layer becomes
Hyperbolic Direct Concat Given a set of hyperbolic points
Lorentz Convolutional Layer Convolution layer is a matrix multiplication between a linearized kernel and a concatenation of the values in its receptive field. Then we generalize this definition by replacing Euclidean operators with hyperbolic counterparts. Given hyperbolic input feature map
The Lorentz convolutional layer is
in which
Parallel Tramsport on Lorentz Model The parallel transport operation
Lorentzian Centroid The weighted centroid with respect to the squared Lorentzian distance, which solves
Lorentz Batch Normalization We generalize batch normalization to the Lorentz model using the Lorentzian centroid and the parallel transport operation for re-centering, and the Fréchet variance and straight geodesics at the origin’s tangent space for re-scaling.
Re-centering Usually Fréchet mean is used, which minimizes the expected distance between set of points in a metric space. However it requires iterative computation, which significantly slows down training process. We use closed form of centeroid w.r.t. Lorentzian distance.
In batch norm, weights are equally assigned, i.e.
Re-scaling We rely on Fréchet variance defined as expected squared Lorentzian distance between points and mean, given by
Lorentz MLR Classifier
Hyperplane in the Lorentz Model
The hyperplane in the Lorentz model is defined by a geodestic that results from the intersection of an
As
in which
where
Theorem. Minimum Hyperbolic Distance
Given
and , the minimum hyperbolic distance from a point to the hyperplane is given by
Proof. Note that the distance from point
where
Finally we insert reparametrization of
Theorem. Output Logit of the Lorentz MLR Given parameters
and , the Lorentz MLR's output logit corresponding to class and input is given by
Lorentz Residual Connection and Activation we add the vector’s space components and concatenate a corresponding time component.
Non-linear Activation
We propose a simpler operation for the Lorentz model by applying the activation function to the space component and concatenating a time component.