All Projects → czczup → Urbanregionfunctionclassification

czczup / Urbanregionfunctionclassification

第五届百度西安交大大数据竞赛 城市区域功能分类 Baseline

Projects that are alternatives of or similar to Urbanregionfunctionclassification

Selective Joint Fine Tuning
Codes and models for the CVPR 2017 spotlight paper "Borrowing Treasures from the Wealthy: Deep Transfer Learning through Selective Joint Fine-tuning".
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Mit Deep Learning
Tutorials, assignments, and competitions for MIT Deep Learning related courses.
Stars: ✭ 8,912 (+11782.67%)
Mutual labels:  jupyter-notebook
Deetctionupperbound
Code for calculating the upper bound AP in object detection
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Mit 6.s094
MIT-6.S094: Deep Learning for Self-Driving Cars Assignments solutions
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Lpproj
Scikit-learn compatible Locality Preserving Projections in Python
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Mml Book.github.io
Companion webpage to the book "Mathematics For Machine Learning"
Stars: ✭ 9,488 (+12550.67%)
Mutual labels:  jupyter-notebook
Ios Coreml Mnist
Real-time Number Recognition using Apple's CoreML 2.0 and MNIST -
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Practicalsessions2019
Materials for the practical sessions at EEML2019
Stars: ✭ 75 (+0%)
Mutual labels:  jupyter-notebook
Pygslib
GSLIB fortran code wrapped into python
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Smartnoise Samples
Code samples and documentation for SmartNoise differential privacy tools
Stars: ✭ 75 (+0%)
Mutual labels:  jupyter-notebook
Ml Examples
some machine learning examples
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Pug classifier
Deep Learning for Pugs
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Pytorch Book
PyTorch tutorials and fun projects including neural talk, neural style, poem writing, anime generation (《深度学习框架PyTorch:入门与实战》)
Stars: ✭ 9,546 (+12628%)
Mutual labels:  jupyter-notebook
Squeezenet Neural Style Pytorch
Lightweight implementation of neural style in Pytorch with SqueezeNet
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Piaic islamabad batch3
Stars: ✭ 75 (+0%)
Mutual labels:  jupyter-notebook
Google earth engine notebook
A walkthrough of some Google Earth Engine Features, as well as using the data in TensorFlow
Stars: ✭ 74 (-1.33%)
Mutual labels:  jupyter-notebook
Nlp Tutorial
Natural Language Processing Tutorial for Deep Learning Researchers
Stars: ✭ 9,895 (+13093.33%)
Mutual labels:  jupyter-notebook
Weibosentiment
基于各种机器学习和深度学习的中文微博情感分析
Stars: ✭ 75 (+0%)
Mutual labels:  jupyter-notebook
Computer vision project
计算机视觉项目实战
Stars: ✭ 75 (+0%)
Mutual labels:  jupyter-notebook
Covid 19
Data analysis and visualizations of daily COVID cases report
Stars: ✭ 75 (+0%)
Mutual labels:  jupyter-notebook

城市区域功能分类

简介

模型结构图

使用遥感图像和访问数据两个模态,在特征层进行融合,大概能拿到0.57的准确率。

tensorboard

快速起步

1.1 依赖环境

tensorflow-gpu==1.8
opencv-python
pandas 

1.2 数据准备

将数据放在data文件夹下,如下所示:

  • data/test_image/test/xxxxxx.jpg
  • data/test_visit/test/xxxxxx.txt
  • data/train_image/train/00x/xxxxxx_00x.jpg
  • data/train_visit/xxxxxx_00x.txt

把压缩文件放在data文件夹里直接解压应该就是上面这样。

我把给的训练集划分了一部分当验证集,具体过程看check_data.ipynb。

划分后的文件名记录在data/train.txt和data/valid.txt中。

1.3 数据转换

把visit数据转换为7x26x24的矩阵,这一步耗时比较长,大概要一个小时。

python visit2array.py

转换后的数据存储在:

  • data/npy/train_visit
  • data/npy/test_visit

1.4 生成tfrecord

python tfrecord.py

生成的tfrecord存储在:

  • data/tfrecord/train.tfrecord
  • data/tfrecord/valid.tfrecord

备注:由于这里直接加载了所有数据,大约要占用5G内存。

1.5 训练

python train.py

为了调参方便,每组实验存在不同的文件夹里。 需要输入显卡的编号和文件夹名称,比如:

device id: 0
dir id: 1001

查看tensorboard:

cd model/
tensorboard --logdir=./

1.6 测试

python test.py

测试完成后在result文件夹中生成结果。

实测分数

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].