Skip to content

Latest commit

 

History

History
18 lines (14 loc) · 556 Bytes

README.md

File metadata and controls

18 lines (14 loc) · 556 Bytes

TSNE

Visulize the Intermidiate features of CNN models using TSNE

In this repo, we use ResNet18 model to classify CIFAR-10 dataset. Then we apply the PCA+TSNE to visualize the features of last FC layer.

Demo of testing data

Img

Get started

Clone this repo and run the code

$ git clone https://github.com/13952522076/TSNE.git
$ cd TSNE
$ CUDA_VISIBLE_DEVICES=0 python3 cifar.py  

Optional paramters: '--number': Random select N exmaples for plotting