Peek-a-Boo: What (More) is Disguised in a Randomly Weighted Neural Network, and How to Find It Efficiently
This repository is the official implementation for the following paper Analytic-LISTA networks proposed in the following paper:
“Peek-a-Boo: What (More) is Disguised in a Randomly Weighted Neural Network, and How to Find It Efficiently” by Xiaohan Chen, Jason Zhang and Zhangyang Wang from the VITA Research Group.
The code implements the Peek-a-Boo (PaB) algorithm for various convolutional networks and is tested in Linux environment with Python: 3.7.2, PyTorch 1.7.0+.
Getting Started
Dependency
Prerequisites
- Python 3.7+
- PyTorch 1.7.0+
- tqdm
Data Preparation
To run ImageNet experiments, download and extract ImageNet train and val images from http://image-net.org/. The directory structure is the standard layout for the torchvision datasets.ImageFolder, and the training and validation data is expected to be in the train/ folder and val/