Japanene OCR Dataset
Korean OCR Dataset
Quantity：15000 images
Format：txt
Quantity：15000 images
Format：txt
OCR Dataset
Japanese and Korean ocr data set, image scenes covering menus, screenshots, books, tickets and road signs.
Quote >
Human Matting Dataset
Semantic Segmentation
Quantity：40000 images
Format：png
Quantity：100000 images
Format：json
Human body Dataset
Image scenarios include indoor, outdoor and other punlic scenes, the human body parts can be used for a variety of visual algorithm applications.
Quote >
|
|