Skip to content

Latest commit

 

History

History
104 lines (92 loc) · 4.4 KB

CHANGELOG.md

File metadata and controls

104 lines (92 loc) · 4.4 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[unreleased]

Changed

  • fix bug in unique labels
  • add crop shape and overlap shape to filename when cropping ds

[1.1.2] - 2024-01-30

Changed

  • Moved package building to pyproject.toml

Fixed

  • Fixed Pipy build

[1.1.1] - 2024-01-22

Fixed

  • Try to fix pypy build

[1.1.0] - 2024-01-22

Changed

  • split_train_test now filters samples with too few unique sources to prevent errors in sklearn
  • refactored to_abs and to_rel method for easier readability and better performance

Added

  • retry reading a parquet file without opening as buffer if it fails
  • semantic_segmentation can store files with numeric names
  • new function: split_train_test_multilabel for fairly splitting multilabel datasets
  • added requirements

Fixed

  • vis_and_store can handle multilabel, singlelabel and None correctly now
  • polygon validation fixed. Set required amount of points from 4 to 3.
  • fixed in validate_unique_annos pandas drop() future warning
  • crop_dataset does not produce empty duplicates of crop positions anymore
  • segmentation_to_lost fix bug where some contours of different classes were merged accidentially
  • mask_dataset won't overwrite input dataframe anymore
  • added a serialization case where a columns has empty lists

[1.0.0] - 2023-01-10

Added

  • fixed voc_eval bug where gt_df bboxes weren't handled correctly
  • fixed coco_eval bug - different dataframes having same uids for images and labels now
  • to_coco meethod has arguments predef_img_mapping and predef_lbl_mapping now to hand over predefined label and image uids
  • export method crop_img to lost_ds (lost_ds.crop_img and ds.crop_img formerly lost_ds.DSCropper.crop_img)
  • resolve FutureWarning at transform_bbox_style
  • make crop_dataset compatible with pandas > 1.5.0
  • crop_component indexing is imagewise instead of global
  • improve some filter and transform functions

Breaking changes

  • Add parallel option for multiple functions
  • fix typo in some methods arguments

[0.5.0] - 2022-12-02

Added

  • method split_train_test which allows the stratify option for i.e. classification datasets
  • coco_eval method for detection (mAP, Average Recall)
  • voc_eval method for detection (tp, fp, fn, precision, revall, ap)
  • to_coco improvements and bugfixes
  • voc_score_iou_multiplex method: shifting bbox score and iou_thresholds to find optimal thresolds
  • cropping method return additional column 'crop_position' now
  • added arg 'random_state' for dataset-splitting
  • added arg for optional parallelistaion
  • added color selection for vis and store - can take column now
  • improved detetion metrics

[0.4.0] - 2022-05-01

Changed

  • file_man: Allow other Fsspec filesystems

Added

  • Progress callback for pack_ds
  • Argument fontscale at vis_and_store to enable manually control for textsize
  • Argument cast_others at segmentation_to_lost to allow ignoring unspecified pixel values

[0.3.0] - 2022-04-05

Added

  • Added zip support for pack_ds method

[0.2.0] - 2022-04-04

Added

  • bbox_nms method to directly apply non-max-suppresseion for bboxes
  • Argument path_col in methods to_abs and to_rel to enable explicit column specification to look for image paths
  • Argument mode for method remap_img_path to enable not only replacement but also prepending of root paths
  • Set argument default of inplace of method pack_ds to False
  • Enable visualization of annos without anno_lbl specified

[0.1.0] - 2022-02-09

Changes

  • validate_unique_annos uses data-hash for dropping duplicates now
  • vis_and_store text-boxes won't be outside the image anymore
  • vis_and_store can determine optimal text-size for labeling when passing arg. line_thickness 'auto'.

Added

  • dependencies for lost_ds in requirements.txt + setup.py
  • examples dataset and code-snippets
  • improved function to load datasets at LOSTDataset
  • new function segmentation_to_lost to convert pixelmaps to lost-annotations
  • new function crop_components to crop dataset based on annotations
  • new function vis_semantic_segmentation to color sem-seg. dataset
  • new function to_coco to generate and store coco datasets from LOSTDataset

[0.0.0] - 2021-10-26

Added

  • First version