[2018-04-21 Spring cloud microservice architecture (1): service registration and mutual invocation ]
Recently the company received a new project with many modules. The single-application architecture cannot meet the needs. The company tried to use the micro-service framework and was entangled in spring cloud or dubbo. Another colleague and I were responsible for spring could learning and dubbo learning respectively. In the end, considering that the changes to the old project were too large, spring cloud was not selected in the end.
I plan to write nearly 15 spring cloud blog posts in the near future, including basic usage and source code analysis, this is the first one.
[2018-04-13 Mask RCNN trains its own data set ]
Collect tongue images of colleagues in the company and make your own data set to identify the tongue.
[2018-03-10 [Target Detection] Install the image labeling tool labelme under Windows ]
The tool needed in the company's tongue medical image diagnosis project.
[2018-03-10 [target detection] Detailed explanation of Mask RCNN algorithm ]
After reproducing Mask RCNN, this blog post is a summary of Mask RCNN.
[2018-03-07 [Target Detection] Detailed explanation of Faster RCNN algorithm ] I
came back for the New Year and learned about the RCNN algorithm (fast rcnn / faster rcnn / mask rcnn). This blog post is a summary of faster rcnn.
[2018-01-30 The reason for the 'Dead ReLU Problem' ]
After reading some Chinese blogs about the advantages and disadvantages of activation functions, few people explain the 'Dead ReLU' phenomenon. There is this translation.
[2018-01-25 Understanding of Forward Propagation and Back Propagation in Neural Networks ] I
finished learning the basic algorithms of machine learning last week, and started to learn basic neural networks and convolutional neural networks this week. This article is my own understanding of basic neural networks.
[2018-01-08 Usage of slices in Python and Numpy (such as: X[:,0], X[:,1], etc.) ]
Since the company is preparing to develop new products (medical image recognition) recently, a new department has been established for research and development In the third department, I, the former head of the first R&D department and some employees were transferred to the new department to be responsible for the initial research and development of new products.
Techniques used include: Python【Numpy、Pandas、Matplotlib等】、DL【Tensorflow、caffe】etc.
Algorithms include: logistic regression, linear regression, decision tree, Bayesian.
Currently learning related knowledge.
[2017-12-28 B-tree (B-tree) insertion and deletion operations ]
After reading CurrentHashMap1.7the source code before, the source code that will be looked at next CurrentHashMap1.8. Since the CurrentHashMap1.8source code involves some knowledge of red-black trees, I went back to review the knowledge of AVL trees, B-trees, B+ trees, and red-black trees, so I have this blog post.