![]() ViTSTR: Vision Transformer for Fast and Efficient Scene Text Recognition.MASTER: MASTER: Multi-Aspect Non-local Network for Scene Text Recognition.SAR: Show, Attend and Read:A Simple and Strong Baseline for Irregular Text Recognition.CRNN: An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition.LinkNet: LinkNet: Exploiting Encoder Representations for Efficient Semantic Segmentation.DBNet: Real-time Scene Text Detection with Differentiable Binarization.Models architecturesĬredits where it's due: this repository is implementing, among others, architectures from published research papers. To interpret your model's predictions, you can visualize them interactively as follows: If both options are set to False, the predictor will always fit and return rotated boxes. Otherwise, if assume_straight_pages=False, it will return rotated bounding boxes (potentially with an angle of 0°). Will be converted to straight boxes), you need to pass export_as_straight_boxes=True in the predictor. If you want the predictor to output straight boxes (no matter the orientation of your pages, the final localizations On your page and return straight boxes, which makes it the fastest option. ![]() If you only use straight document pages with straight words (horizontal, same reading direction),Ĭonsider passing assume_straight_boxes=True to the ocr_predictor. Should you use docTR on documents that include rotated pages, or pages with multiple box orientations, # Analyze result = model( doc) Dealing with rotated documents ![]() models import ocr_predictor model = ocr_predictor( pretrained = True)
0 Comments
Leave a Reply. |