Vines' Log

In this article we try to understand ProGAN, which is a key ingredient in StyleGAN. The central idea of ProGAN is to train both generator and discriminator in gradually increasing resolutions, to provide the network the ability to learn lower level structure first and finer details later.

Studying image-to-image translation. Overview of 2017 CVPR paper "Image-to-Image Translation with Conditional Adversarial Networks" and 2017 ICCV paper "Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks".

Overview on common Generative adversarial network (GAN) methods. We walkthrough GAN, DCGAN, WGAN, LSGAN, and SNGAN.

Studying image-to-image translation. Overview of 2018 ECCV paper "Multimodal Unsupervised Image-to-Image Translation".

I want to keep processes running after SSH session disconnects. During ssh into ubuntu, we can use `nohup` and `&` to ignore the input signal.

I want to set up a SSH server that is on a different port on ubuntu.

I have a user profile all setup with all settings the way I want them. I want to copy the entire configured profile to a new user account. With these command, I can have anaconda and python setups in a newly created profile.

CNN is for classification and RCNN is for object detection. The difference between object detection algorithms and classification algorithms is that in detection algorithms, we try to draw a bounding box around the object of interest (localization) to locate it within the image.

ML techniques in Speech and Speaker Recognition. Including HMM, DNN-HMM, GMM-UBM, GMM-SVM, i-vector, x-vector, and i-vector/PLDA.

Some Practice Questions on traditional ML methods. The topics cover SVM, K-means clustering, Gaussian Mixture Modelling (GMM), EM algorithm, and Dimension Reduction approaches.