Japanese OCR Images Data - Images with Transcription

People with their head-shots in various poses and angles.

People with their head-shots in various poses and angles.

People from the UK with their head-shots in various poses and angles.

Dataset containing open-ended questions about images. These questions require an understanding of vision, language and commonsense knowledge to answer.

Cityscapes is a large-scale urban street-scene dataset with stereo video and high-quality pixel-level annotations, built for benchmarking semantic segmentation, instance segmentation, and panoptic scene understanding for autonomous driving and smart-city computer vision.

People with their head-shots in various poses and angles.