涉及论文：
- Graph Attention Convolution for Point Cloud Semantic Segmentation
- Dual Attention Network for Scene Segmentation
在一篇标题包含“Attention”的论文中，你可能会看到以下公式：
a i j , k = e x p ( a ~ i j , k ) ∑ l ∈ N ( i ) e x p ( a ~ i l , k ) a_{ij,k} = \frac{\mathrm{exp}(\tilde{a}_{ij,k})}{\sum_{l \in \mathcal{N}(i)}\mathrm{exp}(\tilde{a}_{il,k})} aij,k=∑l∈N(i)exp(a~il,k)exp(a~ij,k)
h i ′ = ∑ j ∈ N ( i ) a i j ∗ M g ( h j ) + b i h_{i}^{'} = \sum_{j \in \mathcal{N}(i)}a_{ij}*M_{g}(h_{j})+b_{i} hi′=j∈N(i)∑aij∗Mg(hj)+bi
或者
s j i = e x p ( B i ⋅ C j ) ∑ i = 1 N e x p ( B i ⋅ C j ) s_{ji} = \frac{\mathrm{exp}(B_{i} \cdot C_{j})}{\sum_{i = 1}^{N}\mathrm{exp}(B_{i} \cdot C_{j})} sji=∑i=1Nexp(Bi⋅Cj)exp(Bi⋅Cj)
E j = α ∑ i = 1 N ( s j i D i ) + A j E_{j} = \alpha \sum_{i = 1}^{N}(s_{ji}D_{i}) + A_{j} Ej
最低0.47元/天 解锁文章
1万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
