PASCAL VOC - AshokBhat/ml GitHub Wiki
PASCAL VOC(Visual Object Classes) project
- PASCAL (Pattern Analysis, Statistical modelling and ComputAtional Learning)
- Provides standardized image datasets for object class recognition
- Provides a common set of tools for accessing the datasets and annotations
- Enables evaluation and comparison of different methods
- Ran challenges evaluating performance on object class recognition (from 2005-2012, now finished)
PASCAL VOC data sets
VOC2007
- Data from the PASCAL Visual Object Classes Challenge 2007
- 9963 images in total
- 20 different object classes
- 24640 annotated objects in total
VOC2012
- Data from the PASCAL Visual Object Classes Challenge 2012
- 11540 images in total
- 20 different object classes
- 27450 annotated objects in total
Other datasets
- [ImageNet]] ](/AshokBhat/ml/wiki/[[MNIST) | CIFAR
- [PASCAL VOC]] ](/AshokBhat/ml/wiki/[[COCO) | MS COCO
See also