Notes of "deep learning practice of Web security": Chapter 8 harassment message recognition

This chapter mainly takes SMS Spam Collection data set as an example to introduce the identification technology of harassing SMS. This paper introduces the feature extraction methods used to identify harassing short messages, including word bag and TF-IDF model, vocabulary model, Word2Vec and Doc2Vec model, and introduces the models used and th ...

Posted by neo926 on Thu, 10 Mar 2022 12:44:26 +0100

Win10/Win11 subsystem - wsl2 + Ubuntu 20 04 installation record

preface My last stubbornness - the slogan of not changing windows was defeated by reality. Installing dual systems will affect the part of deep learning that has been done well. Installing virtual machines can't use hardware GPU. I accidentally found that windows has such a thing, and many people are already using it. PS: after the instal ...

Posted by freynolds on Thu, 10 Mar 2022 11:47:46 +0100

Introduction to neural network quantization -- quantitative perception training

This article begins with the official account of "AI little boy". Last article This paper introduces the basic process of post training quantization, and demonstrates the simplest post training quantization algorithm with pytorch. Although post training quantization is easy to operate, most reasoning frameworks provide such off-line ...

Posted by razmon on Thu, 10 Mar 2022 02:08:15 +0100

Hengyuan cloud (GpuShare)_ Medical image segmentation: MT Unet

Our community has new technology sharing partners ๐ŸŽ‰๐ŸŽ‰๐ŸŽ‰a warm welcome ๐Ÿ‘As a qualified Porter, I must do something to express my joy: Handling ~ handling ~ immediate handling ~Article source| Hengyuan cloud communityOriginal address| New hybrid Transformer module (MTM)Original author Dong DongabstractExisting problemsAlthough U-Net has achieved gr ...

Posted by Pyro4816 on Wed, 09 Mar 2022 12:04:00 +0100

Code analysis -- model creation

https://blog.csdn.net/Dear_learner/article/details/122920181 ย  Two elements of building a model: โ— build sub modules (such as convolution layer, pooling layer, activation layer and full connection layer in the network structure); โ— splice sub modules (splice the sub modules in a certain order to finally get the desired network structure). ย  The ...

Posted by rowman on Wed, 09 Mar 2022 03:03:51 +0100

Full convolution network (FCN) practice: using FCN to realize semantic segmentation

Full convolution network (FCN) practice: using FCN to realize semantic segmentation FCN classifies images at pixel level, which solves the problem of semantic segmentation. Unlike the classical CNN, which uses the full connection layer to obtain a fixed length feature vector for classification after the convolution layer (full connection layer ...

Posted by purinkle on Wed, 09 Mar 2022 00:03:14 +0100

yolov5 training data set (ultralytics version)

0 preparation Environment configuration, reference hereDownload an official yolov5s PT model, placed in yolov5 root directory, download link here perhaps herePrepare the dataset in the format shown below. Each picture has its corresponding xml format annotation file. yolov5 โ”œโ”€โ”€ models โ”œโ”€โ”€ runs โ”œโ”€โ”€ utils โ”œโ”€โ”€ data โ”‚ โ”œโ”€โ”€ project01 โ”‚ โ”‚ โ”œโ”€โ”€ ...

Posted by chinstroker on Tue, 08 Mar 2022 16:59:09 +0100

Briefly introduce the use of distributed training DDP in pytorch (combined with examples, quick start)

DDP principle Distributed data parallel (DDP) supports multi machine and multi card distributed training. pytorch native support. This paper briefly summarizes the use of DDP and the test under multi card, and introduces it according to the actual code. voxceleb_trainer: open source voiceprint recognition tool, simple and easy to use, ...

Posted by sendoh07 on Tue, 08 Mar 2022 16:16:02 +0100

Recommended system - Neural Collaborative Filtering(NMF)

preface Thesis address: https://arxiv.org/pdf/1708.05031.pdf?ref=https://codemonkey.link Code address: https://github.com/hexiangnan/neural_collaborative_filtering motivation This is an earlier article. Recommendation systems often use the dot product of vectors to learn the interactive information between vectors, Purpose of vector dot ...

Posted by steveonly20 on Tue, 08 Mar 2022 10:19:51 +0100

[Python + + opencv + Python] license plate extraction, segmentation and recognition

If you want to complete the project, I will put the github link at the end of the article: You can see that the final identification license plate number is min G99999. In fact, I accidentally thought of doing a small project with c + + in the winter of the previous year, and then realized the license plate extraction and segmentation with c+ ...

Posted by ricta on Tue, 08 Mar 2022 05:50:40 +0100