Dataset Groups Activity Stream Total-Text Total-Text is a dataset for word-level arbitrary-shaped English text detection, containing 1,255 images for training and 300 images for testing. BibTex: @dataset{Pengyuan_Lyu_and_Minghui_Liao_and_Cong_Yao_and_Wenhao_Wu_and_Xiang_Bai_2024, abstract = {Total-Text is a dataset for word-level arbitrary-shaped English text detection, containing 1,255 images for training and 300 images for testing.}, author = {Pengyuan Lyu and Minghui Liao and Cong Yao and Wenhao Wu and Xiang Bai}, doi = {10.57702/5h8rjl9j}, institution = {No Organization}, keyword = {'Detection', 'Scene Text', 'Scene Text Detection', 'Text Detection', 'Text Recognition', 'Word-level Text', 'dataset', 'scene text', 'text detection', 'text recognition'}, month = {dec}, publisher = {TIB}, title = {Total-Text}, url = {https://service.tib.eu/ldmservice/dataset/total-text}, year = {2024} }
