MMOCR Release v1.0.0rc6

Highlights

1. Two new models, ABCNet v2 (inference only) and SPTS are added to projects/ folder.
2. Announcing Inferencer, a unified inference interface in OpenMMLab for everyone's easy access and quick inference with all the pre-trained weights. Docs
3. Users can use test-time augmentation for text recognition tasks. Docs
4. Support batch augmentation through BatchAugSampler, which is a technique used in SPTS.
5. Dataset Preparer has been refactored to allow more flexible configurations. Besides, users are now able to prepare text recognition datasets in LMDB formats. Docs
6. Some textspotting datasets have been revised to enhance the correctness and consistency with the common practice.
7. Potential spurious warnings from shapely have been eliminated.

Dependency

This version requires MMEngine >= 0.6.0, MMCV >= 2.0.0rc4 and MMDet >= 3.0.0rc5.

New Features & Enhancements

• Discard deprecated lmdb dataset format and only support img+label now by @gaotongxiao in #1681
• abcnetv2 inference by @Harold-lkk in #1657
• Add RepeatAugSampler by @gaotongxiao in #1678
• SPTS by @gaotongxiao in #1696
• Refactor Inferencers by @gaotongxiao in #1608
• Dynamic return type for rescale_polygons by @gaotongxiao in #1702
• Revise upstream version limit by @gaotongxiao in #1703
• TextRecogCropConverter add crop with opencv warpPersepective function by @KevinNuNu in #1667
• change cudnn benchmark to false by @Harold-lkk in #1705
• Add ST-pretrained DB-series models and logs by @gaotongxiao in #1635
• Only keep meta and state_dict when publish model by @Harold-lkk in #1729
• Rec TTA by @Harold-lkk in #1401
• Speedup formatting by replacing np.transpose with torch… by @gaotongxiao in #1719
• Support auto import modules from registry. by @Harold-lkk in #1731
• Support batch visualization & dumping in Inferencer by @gaotongxiao in #1722
• add a new argument font_properties to set a specific font file in order to draw Chinese characters properly by @KevinNuNu in #1709
• Refactor data converter and gather by @Harold-lkk in #1707
• Support batch augmentation through BatchAugSampler by @gaotongxiao in #1757
• Put all registry into registry.py by @Harold-lkk in #1760
• train by @gaotongxiao in #1756
• configs for regression benchmark by @gaotongxiao in #1755
• Support lmdb format in Dataset Preparer by @gaotongxiao in #1762

Docs

• update the link of DBNet by @AllentDan in #1672
• Add notice for default branch switching by @gaotongxiao in #1693
• docs: Add twitter discord medium youtube link by @vansin in #1724
• Remove unsupported datasets in docs by @gaotongxiao in #1670

Bug Fixes

• Update dockerfile by @gaotongxiao in #1671
• Explicitly create np object array for compatibility by @gaotongxiao in #1691
• Fix a minor error in docstring by @Mountchicken in #1685
• Fix lint by @triple-Mu in #1694
• Fix LoadOCRAnnotation ut by @Harold-lkk in #1695
• Fix isort pre-commit error by @KevinNuNu in #1697
• Update owners by @xinke-wang in #1699
• Detect intersection before using shapley.intersection to eliminate spurious warnings by @gaotongxiao in #1710
• Fix some inferencer bugs by @gaotongxiao in #1706
• Fix textocr ignore flag by @xinke-wang in #1712
• Add missing softmax in ASTER forward_test by @Mountchicken in #1718
• Fix head in readme by @vansin in #1727
• Fix some browse dataset script bugs and draw textdet gt instance with ignore flags by @KevinNuNu in #1701
• icdar textrecog ann parser skip data with ignore flag by @KevinNuNu in #1708
• bezier_to_polygon -> bezier2polygon by @double22a in #1739
• Fix docs recog CharMetric P/R error definition by @KevinNuNu in #1740
• Remove outdated resources in demo/ by @gaotongxiao in #1747
• Fix wrong ic13 textspotting split data; add lexicons to ic13, ic15 and totaltext by @gaotongxiao in #1758
• SPTS readme by @gaotongxiao in #1761

New Contributors

• @triple-Mu made their first contribution in #1694
• @double22a made their first contribution in #1739

Full Changelog: v1.0.0rc5...v1.0.0rc6