SHDA23YL

Data

The data zip file contains 4 subfolders corresponding different regions. California, Illinois, North Dakota, Nebraska. Each subfolder has two subfolders one for Aerial images (Aer), another for Digital Eelevation Model (DEM).

Usage

Create a new conda environment:

conda create --name culvert python=3.9
activate culvert       # Windows
conda activate culvert # Linux

Clone this repo:

git clone https://github.com/SHUs-Lab/SHDA23YL.git
cd SHDA23YL

Data Preparation

1. Unzip ClippedSample_5Areas.zip
2. Run DataRead.py
3. Run VICalculation.py

Data Merge

step0.merge_data.ipynb is for merging the datasets and labels from 4 different regions.

Neural Architecture Search with NNI

step1.NNI.ipynb is for NNI trying different model configurations with ResNet.

Latency Prediction

step2.nn-Meter.ipynb is for latency prediction using nn-Meter.

Pareto Front Analysis

step3.pareto_front.ipynb is for Pareto front analysis with three objectives.

1. Maximize accuracy
2. Minimize latency
3. Minimize model size