https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification

Winning solution scripts
https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification

kaggle mxnet

Last synced: 2 months ago
JSON representation

Winning solution scripts

• Host: GitHub
• URL: https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification
• Owner: u1234x1234
• Created: 2016-04-28T17:39:14.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2017-04-22T12:28:50.000Z (over 7 years ago)
• Last Synced: 2024-08-01T22:41:41.865Z (5 months ago)
• Topics: kaggle, mxnet
• Language: Python
• Size: 38.1 KB
• Stars: 54
• Watchers: 7
• Forks: 23
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• Awesome-MXNet - Yelp Restaurant Photo Classifacation

README

        
# [Kaggle Yelp Restaurant Photo Classification](https://www.kaggle.com/c/yelp-restaurant-photo-classification)

Description of the my approach to this problem was published on the [Kaggle's blog](http://blog.kaggle.com/2016/04/28/yelp-restaurant-photo-classification-winners-interview-1st-place-dmitrii-tsybulevskii/).  

There is no click-get-submission-file script here, only set of the small scripts for different parts of the final solution.

[/model/](https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification/tree/master/model/21k)  

folder contains photo-level feature extraction scripts with different pretrained netoworks/layers.

[compress.py](https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification/blob/master/compress.py)  

Photo-level feature preprocessing: normalization or/and PCA transformation

[fisher.cpp](https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification/blob/master/cpp/feature_extractor/fisher.cpp)  

Fisher Vectors computation, input: features generated by compress.py

[vlad.cpp](https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification/blob/master/cpp/feature_extractor/vlad.cpp)  

VLAD descriptor computation, input: features generated by compress.py

[pool.py](https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification/blob/master/pool.py)  

Business-level feature extraction. Input: features generated by compress.py in case of feature averaging, or from fisher.cpp/vlad.cpp in case of FV/VLAD features.  

[predict_test.py](https://github.com/u1234x1234/kaggle-yelp-restaurant-photo-classification/blob/master/predict_test.py)  

Model training/prediction/submission file generation.  

Input: features generated by different execution of pool.py script.