This article summarizes the important knowledge points related to deep learning. It is presented to you through long graphs and PDFs. You are welcome to download PM.
This article is common to 10 CNNThe visualization of the architecture helps you fully understand the evolution of the CNN architecture.
Their research and some other work during the same period have shown that these models can also be widely used to model the relationship between visual basic elements, including objects and objects, between objects and pixels, and between pixels and pixels, especially in Modeling the relationship between pixels and pixels can complement the convolution operation, and even hope to replace the convolution operation to achieve the most basic image feature extraction.
Deep learning cannot make causal reasoning, and graph model (GNN) is one of the solutions. Professor Sun Maosong of Tsinghua University published a review paper, comprehensively expounded GNN and its methods and applications, and proposed a unified representation that can characterize the propagation steps in various GNN models. In the text chart, it is recommended to print the high-definition plastic stickers for reference.
CNNVery good at classifying out-of-order images, but humans are not. In this article, the authors show why the most advanced deep neural networks still recognize garbled images well, and the reasons for this help reveal that DNN uses an unexpectedly simple strategy to classify natural images.
Recently, Shannon Technology published a research and proposed a Chinese glyph vector Glyce. The research is based on the evolution of Chinese characters, using a variety of Chinese characters and various writing styles, designed for Chinese pictographic characters. CNN Architecture - Tian Character Bank CNN. Glyce in 13 (almost all) Chinese NLP The task has achieved the best performance at the moment.