Clothes classification (50 categories) and auto-tagging (1000 attributes) with Keras based on Tensorflow
Run:
python clothes_classifier.py 2>&1 | tee ./result.log
The former is a single-label classification problem and the latter is a multi-label classification problem.
Dataset: DeepFashion Attribute Prediction
-
Number: 289147 images in total
-
Distribution of the Clothes Category:
category_name number Anorak 160 Blazer 7495 Blouse 24557 Bomber 309 Button-Down 330 Cardigan 13311 Flannel 324 Halter 17 Henley 716 Hoodie 4048 Jacket 10467 Jersey 748 Parka 676 Peacoat 97 Poncho 791 Sweater 13123 Tank 15429 Tee 36887 Top 10078 Turtleneck 146 Capris 77 Chinos 527 Culottes 486 Cutoffs 1669 Gauchos 49 Jeans 7076 Jeggings 594 Jodhpurs 45 Joggers 4416 Leggings 5013 Sarong 32 Shorts 19666 Skirt 14773 Sweatpants 3048 Sweatshorts 1106 Trunks 386 Caftan 54 Cape 0 Coat 2120 Coverup 17 Dress 72083 Jumpsuit 6153 Kaftan 126 Kimono 2294 Nightdress 0 Onesie 70 Robe 150 Romper 7408 Shirtdress 0 Sundress 0 -
Distribution of the Clothes Attributes (Omit)
- 10.19 19/50-category classification: 0.3268 F1-score on validation set
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent