摘要： Read Data: 0826 Publication: CVPR 2020 Title: MixNMatch: Multifactor Disentanglement and Encoding for Conditional Image Generation Participants: Yuhen 阅读全文

摘要： Read Data: 0825 Publication: CVPR 2021 Title: Verifiability and Predictability: Interpreting Utilities of Network Architectures for Point Cloud Proces 阅读全文

摘要： Read Data: 0824 Publication: CVPR 2021 Title: Where and What? Examining Interpretable Disentangled Representations Participants: Xinqi Zhu, Chang Xu, 阅读全文

摘要： Read Data: 0705 Publication: ICCV 2019 Title: ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks Participant 阅读全文

摘要： Read Data: 0629 Publication: ECCV 2014 Title: Visualizing and Understanding Convolutional Networks Participants: Matthew D. Zeiler ; Rob Fergus Aim: 为 阅读全文

摘要： Read Data: 0628 Publication: Title: Dynamic Head: Unifying Object Detection Heads with Attentions Participants: Xiyang Dai Yinpeng Chen Bin Xiao Dongd 阅读全文

摘要： 反卷积作用 反卷积即之前的转置卷积，能够起到上采样作用。 操作过程 整体的流程如下图所示。右边是正常网络，左边是反卷积过程。其中，反池化和反卷积是重点。 详细内容见笔记 理解 随着网络的由浅及深，提取的特征由抽象到具体，由颜色、边缘到关键性特征。 总结 反卷积算法，不仅用于可视化，而且能够应用于网络 阅读全文

摘要： 转置卷积 转置卷积又称做反卷积，出自《A guide to convolution arithmetic for deeplearning》论文。可以起到上采样以及可视化作用。 1.转置卷积的操作 相关笔记整理如下图。 2.转置卷积的实现代码 import numpy as np import ma 阅读全文

摘要： Read Data: 0620 Publication: 网络 Title: RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition Participants: Xiaohan 阅读全文

摘要： Read Data: 0618 Publication: CVPR 2021 Title: Stand-Alone Self-Attention in Vision Models Participants: Prajit Ramachandran，Niki Parma，Ashish Vaswani， 阅读全文