1. Introduction In this post, I write about reviews of some ResNet concerned paper, which includes ResNet, ResNeXt, SENet, SKNet, and ResNeSt. Since the publication of VGG net, lots of efforts are made to develop more efficient and more modular network backbone. The most popular 2 models are Inception and... [Read More]

1. Introduction In this post, I would like to compare the loss functions used in different one-shot object detection methods, YOLO, SSD, and RetinaNet. One-shot object detection methods train the model on more than thousands grids with different scale, but the number of objects in one image is much less.... [Read More]

1. Introduction In this post, I resume the development of Inception network from V1 to V4. The main purpose of this post is to clearly state the development of design of Inception network. For better understanding of the history, I list the time of the publication of the 4 paper... [Read More]

1. Introduction I started coding 15 years ago. During this 15 years, I spent most of the time in the academic domain. No one forces me to follow the coding style. Although I learned some small skills fo coding sometimes, it’s not systematical. Now I would like to study the... [Read More]

1. Introduction To reproduce a deep neural network model, I have to get clear of three important elements: Network architeture Loss function Training practices [Read More]

1. Introduction In this post, I’d like to review the 3 paper of YOLO. The main purpose is to understand the design of the YOLO and how the authors try to improve YOLO. For the details of implementation, such as learning rate and training tricks, please read the experiments parts... [Read More]