Vision

Image, document and video data for African contexts and scripts. These tasks share the same capture, preprocessing and annotation groundwork.

Shared across image & video data

• Sourcing & rights for images and video
• Capture / scan standards (resolution, format, lighting)
• Preprocessing (deskew, denoise, normalization, encoding)
• Annotation (labels, bounding boxes, segmentation, temporal spans)
• Annotation tools & inter-annotator agreement
• Metadata, dataset formatting & splits
• Licensing & likeness / consent

Tasks

• Image data – (classification, object detection, segmentation)
• OCR & document AI – (printed & handwritten text, layout, document understanding)
• Sign language, gesture & video – (recognition, glossing, sign-language translation)