Skip to content

ZwX1616/Open-Image-challenge-2018

Repository files navigation

Open-Image-challenge-2018

my solution to Google's Open Image Dataset v4 challenge

*for quick setup, read /Note/APPNOTE.py

folders organization: (now supports: mxnet-SSD training (and predicting)) (still working on the voc format conversion)

root
├── sources
│   ├── train_img 
│             ├── 0123456789ABCDEF.jpg //downloaded training images 
│             ├── ...
│   ├── train_xml
│             ├── 0123456789ABCDEF.xml //voc formatted label, generated using gen_xml in /root/dataset_tools
│             ├── ...
│   ├── train_dummy
│             ├── 0123456789ABCDEF.txt //only exists if the JPG's orientation is portrait
│             ├── ...
│   ├── test_img 
│             ├── 0123456789ABCDEF.jpg //downloaded testing images 
│             ├── ...
│   ├── valid_img 
│             ├── 0123456789ABCDEF.jpg //downloaded val images, optional
│             ├── ...
│   ├── sets 
│             ├── series
│                       ├── cls_series_1.txt //subset classes, generated using split_train in /root/dataset_tools
│                       ├── ...
│             ├── train_series_1.txt //voc formatted subset image indexs, generated using split_train in /root/dataset_tools
│             ├── ...
│   └── rec_files
│             ├── train_series_1.lst //mxnet formatted subset label, generated using split_train in /root/dataset_tools
│             ├── train_series_1.rec //mxnet formatted subset image+label pack, generated using mxrec_train in /root/dataset_tools
│             ├── train_series_1.idx //mxnet formatted indexs, generated together with the rec file
│             ├── ...
├── train
│        ├── label
│                ├── train-annotations-bbox.csv
│                ├── validation-annotations-bbox.csv
│                ├── test-annotations-bbox.csv
│                ├── class-descriptions-boxable.csv
│                ├── challenge-2018-class-descriptions-500.csv
│                └── challenge-2018-attributes-description.csv
│        ├── relation
│                ├── ... (some vrd csv)
├── dataset_tools
│   └── <dataset_tools>
│
├── detectors
│   └── <detector models>
│
├── Note
│   └── APPNOTE.py
│
├── outputs
│   └── <training result folders>
│ 
├── predictions
│   ├── <test set prediction results for each series>
│   └── submission.csv
│
└── train_distn_sorted.txt //class names sorted by number of samples

(under construction...)

About

my solution to Google's Open Image Dataset v4 challenge

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published