Releases: Kitware/nrtk-explorer
v0.7.0
v0.7.0 (2024-12-17)
Chore
- chore(ci): on push to release, run semantic-release
Automatically merge release commit back to main.
Delete sync_branches workflow which was erroneously merging
back to main from release before semantic-release was run. (b5aa743)
- chore(pyproject.toml): limit nrtk to 0.16.0 (
c8285cb)
Feature
- feat(core): add Engine kwargs alternative to cli args (
2eb3d0d)
Fix
- fix(dataset): load local hugging face dataset (
9ca54f6)
Unknown
- Merge pull request #160 from Kitware/main
cut 0.7.0 (6e6ab8a)
- Merge pull request #158 from Kitware/kwargs-config
Engine kwargs as alternative to CLI args (ed1e971)
- Merge pull request #154 from Kitware/release
0.6.0 (a0dc46e)
v0.6.0
v0.6.0 (2024-12-12)
Chore
- chore(pyproject.toml): set trame-annotations min to 0.4.0 (
b45bf2c)
Documentation
- docs(nebari): describe and test nebari setup (
227b65d)
Feature
-
feat: add confidence score threshold (
b75b50c) -
feat(stateful_annotations): put annotation score on state (
e2366a8) -
feat(coco_utils): add scoring for classification model (
32660f6) -
feat(stateful_annotations): support non bbox predictions (
be88218) -
feat: support classification datasets (
5428c1c) -
feat(transforms): add ability to add multiple transforms (
3d58599) -
feat(dataset): add download CLI argument for HF datasets
Defaulting to streaming download for HF datasets, but now you can specify
--download to download the dataset to disk before loading it. (f3a84e1)
-
feat(dataset): add streaming option to HF datasets (
1818745) -
feat(dataset): expand HF dataset splits as new datasets (
c613029) -
feat(dataset): load hugging face datasets
Wraps a HF dataset into our Coco like dataset API. (3344920)
-
feat(transforms): add CLI arg for inference models (
9f4a5e0) -
feat(ImageDetection): use component from trame-annotations (
c4cecc6)
Fix
- fix(pyproject): avoid broken pybsm version
Also update scikit-learn
- fix(images): avoid OSError - to many open files
-
fix(dataset): add types (
d24e9e5) -
fix(dataset): add get_image to KWCocoDataset (
389403d) -
fix(dataset): support list/row first datasets (
6c9567b) -
fix(dataset): transforms require RGB mode images (
5c176d7) -
fix(dataset): less picky about dataset feature shape (
9983b21) -
fix(dataset): support more HF datasets
-
Fix converting CMYK image formats to PNG.
-
Remove use of get_image_fpath in app. (
261191e)
Performance
-
perf(dataset): convert image mode to RGB on HF dataset load (
752b56f) -
perf(dataset): loop through dataset once (
a451b97)
Refactor
-
refactor: add annotations lib module (
f3b76c4) -
refactor: rename object_detection_model to inference_model (
ca3f7b4) -
refactor(coco_utils): make normalized annotation struct (
ae6279e) -
refactor(transforms): move score logic to coco_utils (
b61bd34) -
refactor(images): move img.mode("RBG") to image loader (
be817b9) -
refactor(dataset): simplify extract_labels (
9d7c2ee) -
refactor(dataset): share common code between dataset classes (
6b6b381) -
refactor(dataset): rename JsonDataset (
7d4022b) -
refactor(transforms): fix spelling visibile -> visible (
d2fd39d)
Unknown
- Merge pull request #153 from Kitware/main
Cut PiPy 0.6.0 release (a1df9a5)
v0.5.0
v0.5.0 (2024-10-21)
Ci
- ci: type checking (
51110be)
Documentation
-
docs(sphinx): Precommit hooks (
d80270c) -
docs(sphinx): Add DEVELOPMENT instructions. (
115b272) -
docs(sphinx): Add sphinx auto-api docs. (
4a4d88f) -
docs(README): add back installation section (
5b019a6) -
docs(README): format style (
79aec9a) -
docs(README): add release section, screenshot, remove install section (
c5df7e7)
Feature
- feat(perturber): add support for loading more perturber via YAML definition (
a58ac87)
Fix
- fix(trame): ensure newer version of trame to suport used API
- fix(nrtk): update arguments to PybsmPerturber
New version of nrtk broke old sensor scenerio parameter factory we had. (28b3183)
- fix(deps): add nrtk[headless] extra to fix runtime error
nrkt put opencv-python under an extra in 0.12.0 https://github.com/Kitware/nrtk/releases/tag/v0.12.0 (d664ad7)
- fix(transforms): on dataset change stop processing
Fixes key error accessing images in new dataset with old dataset image ids. (9fb545e)
- fix(debounce): only await coroutines
Refactor
- refactor(image-list): reuse AbstractElement state+ctrl
Guard against double call to init_visibile_columns (0d42b9e)
-
refactor(image-list): move in column visibility logic (
51b9fdf) -
refactor(image-list): move image_list_ids logic to ImageList class (
538da8d) -
refactor(image-list): move pagnation state updates to class (
c2dd051) -
refactor(image-list): use TrameApp on ImageList for change decorators (
88d446f)
Unknown
- Merge pull request #139 from Kitware/fix-dep-version
fix(trame): ensure newer version of trame to suport used API (1feffc5)
- Merge pull request #137 from Erotemic/docs
Add documentation (9175aa7)
- Merge pull request #133 from Kitware/yaml-perturbators
Yaml perturbators (58bb073)
- Merge pull request #129 from Kitware/release
0.4.0 (ca2baab)
- Merge release into main (
1c9359e)
v0.4.0
v0.4.0 (2024-10-08)
Chore
- chore(deps): update build requirement from <0.10.0 to <1.3.0
Updates the requirements on build to permit the latest version.
updated-dependencies:
- dependency-name: build
dependency-type: direct:production
...
Signed-off-by: dependabot[bot] <[email protected]> (82bf3c9)
- chore(deps-dev): bump vite from 4.5.3 to 4.5.5 in /vue-components
Bumps vite from 4.5.3 to 4.5.5.
updated-dependencies:
- dependency-name: vite
dependency-type: direct:development
...
Signed-off-by: dependabot[bot] <[email protected]> (d103f97)
- chore(deps): bump rollup from 3.29.4 to 3.29.5 in /vue-components
Bumps rollup from 3.29.4 to 3.29.5.
updated-dependencies:
- dependency-name: rollup
dependency-type: indirect
...
Signed-off-by: dependabot[bot] <[email protected]> (a6932ae)
-
chore(docker): add docker deploy definition (
97db133) -
chore(deps): bump scikit-learn from 1.5.1 to 1.5.2
Bumps scikit-learn from 1.5.1 to 1.5.2.
updated-dependencies:
- dependency-name: scikit-learn
dependency-type: direct:production
update-type: version-update:semver-patch
...
Signed-off-by: dependabot[bot] <[email protected]> (ebace6e)
Documentation
- docs(README): update screenshot (
1a834f9)
Feature
-
feat(core): default to 500 sampled dataset images (
2e763ce) -
feat(ScatterPlot): stay in selection mode after selection (
1edd988) -
feat(ScatterPlot): dark gray for selected points
Light gray for unselected points. Was light blue for unselected points, light gray for selected. (5f482c8)
-
feat(transforms): add 3 more object detection models (
b2f65ce) -
feat(image_list): only show spinner if show annotations is on
or image is loading (a1a3e18)
-
feat(image_list): use normal switch for show annotations (
f527fef) -
feat(image_list): add switch to hide annotations (
12f2971) -
feat(image_list): slider for image size (
187bd66) -
feat(filtering): relabel Apply button to Select Images (
59d9da9) -
feat(layout): move category filter to bottom (
b4dec14) -
feat(transforms): toggle switch to disable object detection (
b6cb739) -
feat(transforms): toggle component to disable transforms (
0d79fca) -
feat(transforms): add disable transform option (
5cd2601)
Fix
-
fix(ImageDetection): stop tooltip overflow with fudge to center (
437b6a5) -
fix(core): dataset path does not overflow select dropdown (
b33b975) -
fix(ScatterPlot): reactive color map control and perf
Big performance improvements for many point (6197f16)
- fix(embeddings): disable transforms switch hides points
-
fix(images): dont remove image from cache for embeddings (
e0afacd) -
fix(transforms): turn on transform enabled switch on apply button (
c821022) -
fix(dep): properly describe expected version (
d3c060b) -
fix(network): ensure network completion before heavy work (
26c73c0) -
fix(ui): use class components (
5fafd29) -
fix(embeddings): dont send empty lists to dim reducer (
ce24374) -
fix(embedding): attempt to revive standalone mode (
a68d917) -
fix(embedding): standalone mode works again (
f577e99) -
fix(project): cleanup dependency structure (
6f64c10)
Performance
- perf(core): add debounce to sample size slider
- perf(images): add cache backed get_stateful_image funcs
Sometimes we don't need the image for visualization and don't
need to put it on the trame state. (356c78a)
Refactor
-
refactor(annotations): reuse LruCache for annotations (
3f5f704) -
refactor(images): move module level funcs to class (
3d306e6) -
refactor(transforms): extract detection and transform enable/disable logic (
a6252fa)
Unknown
-
All sampled images in image list (#88)
-
feat: show all images in list
-
feat(embeddings): selection of points filters list
-
refactor(object_detector): remove images_manager dependency
-
refactor: move image modules into images folder
-
refactor: use BufferCache for images
-
refactor: use BufferCache for annotations
-
refactor: clean dead code
-
fix(embeddings): add transformed img point as computed
-
chore(ci): run tests without depending on linters
-
fix(image_list): respect client side sorting and filtering
-
fix(ScatterPlot): correct is_transformed in hover
-
fix(image_list): paginate grid view
Grid view does not do virtual scrolling
-
perf(object_detector): reuse last successful batch size
-
refactor: remove images_manager
-
fix: flush transformed images to state before detection
-
fix(images): actually call on_clear callback in BufferCache
-
feat(image_list): hide dependant columns when transforms disabled
-
refactor(images): use lru_cache for image functions
-
refactor(ScatterPlot): rename props to use points
ScatterPlot could be used for non images data.
-
refactor: reorder images.py and doc object_detector
-
refactor(annotations): move logic from images.py (
9850cf6)
v0.3.2
v0.3.2 (2024-08-30)
Ci
-
ci: add noop release check (
1a48185) -
ci: set make_release as manual job (
0fb5fa2) -
ci: add sync job (
7962ee3)
Fix
-
fix(dataset): set kwcoco optional backend (
7400518) -
fix(transforms): parametrize pybsm darkCurrentFromDensity call (
af34dbb)
Unknown
- Merge pull request #101 from Kitware/main
Bring master changes (9c87f92)
v0.3.1
v0.3.0
v0.3.0 (2024-08-07)
Chore
- chore(deps): bump scikit-learn from 1.5.0 to 1.5.1
Bumps scikit-learn from 1.5.0 to 1.5.1.
updated-dependencies:
- dependency-name: scikit-learn
dependency-type: direct:production
update-type: version-update:semver-patch
...
Signed-off-by: dependabot[bot] <[email protected]> (95001d5)
-
chore(coco_utils): avoid circular import for tests (
551fe22) -
chore: update npm packages (
b8b245f) -
chore(deps-dev): bump vite from 4.5.2 to 4.5.3 in /vue-components
Bumps vite from 4.5.2 to 4.5.3.
updated-dependencies:
- dependency-name: vite
dependency-type: direct:development
...
Signed-off-by: dependabot[bot] <[email protected]> (e57e3b7)
- chore(deps): bump scikit-learn from 1.4.2 to 1.5.0
Bumps scikit-learn from 1.4.2 to 1.5.0.
updated-dependencies:
- dependency-name: scikit-learn
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <[email protected]> (3099684)
Ci
Feature
-
feat(image_list): add visible columns control (
1c87a6e) -
feat(transforms): async display of ground truth annotations (
6b434ee) -
feat: use kwcoco for coco datasets (
782d6a3) -
feat(dataset): add force_reload parameter (
cfb2b63) -
feat(transforms): always show ground truth
Always compute score against ground truth. (5c6cde6)
-
feat(image_list): replace grid switch with toggle button (
b95031d) -
feat(ImageDetection): only show annotation ID for ground truth
Was showing repetitive category ID for object detection model
annotations. (396e3e4)
- feat: toggle ground truth/predictions annotations
- Toggle ground truth/predictions annotations for source and
transformation images. - Compute COCO score (NRTK) instead of embeddings cartesian distance.
- Add unit tests for the coco_utils.py newly module. (
0ab0b5b)
-
feat(image_list): add search and fullscreen controls (
3196272) -
feat(ImageDetection): keep tooltip visible and hide scrollbars
-
fix img css position to block to avoid following image
content from shifting as image did not contribute to component width. -
refactor canvas context creation. (
d235d14) -
feat(image_list): add embeddings distance to image table (
1d4f6dd) -
feat(image_list): add table panel under tab (
172b256) -
feat(image_list): spinner for loading images (
1c162e0) -
feat(image_list): group original and transformed images (
6f0d803)
Fix
- fix(core): filter image selection by ground truth categories
-
fix(embeddings): hover on original image highlights (
0b7d00b) -
fix(dataset): prefer implicit return types
get_dataset() returns a CocoDataset, not a Dataset. (bc5b0ca)
-
fix: entry name of transforms subapp (
363716d) -
fix(ImageDetection): keep the same color for same labels (
9e0a733) -
fix: dataset.py snake case and remove unused imports (
53b6005) -
fix: show object detection label if no category in dataset (
b4f9b41) -
fix(transforms): compute score for truth to transform correctly (
141bea6) -
fix: score 0 when no predictions, dataset switching bug (
9cb8ab1) -
fix(ImageDetection): keep tooltip from clipping under table
Shifting tooltip if out of window was not enough
when table size was small, and tooltip would clip
under the table footer. (a670d92)
- fix(transforms): resize transformed image for scoring
Annotation similarity scoring requires the transformed image
to be the same size as the original image. (86c0e2e)
- fix(transforms): use class agnostic scorer
Avoids error when Object Detection Model outputs
category that is not in COCO JSON. (d3bccfe)
-
fix(image_meta): add distance to meta type dict (
43a704e) -
fix(pyproject): set min versions for nrtk and timm (
01ce0bb) -
fix(ScatterPlot): after selection, start navigation mode again (
eb65f38) -
fix(embeddings): allow selection of transformed point (
dd1650f) -
fix(ImageDetection): avoid extra bottom padding on hover border (
95235a5) -
fix(ImageDetection): miss aligned annotations for transformed images
When transformed image resolution was different from the original image,
the annotations were not aligned with the image.
-
fix(embeddings): distinguish image kind on hover (
26a670d) -
fix(image-list): show hover box on original or transformed (
e61e66c)
Performance
- perf(transforms): only create transformed images if visible (
e11b422)
Refactor
-
refactor(trame_utils): simplify and doc change_checker (
0ac0975) -
refactor(image-list): add image server
Keeps original image off state
Remove commented code (16a953d)
- refactor(transforms): dont duplicate annotations on context
Stashing them on state once is enough (65e5c8f)
v0.2.2
v0.2.1
v0.2.0
v0.2.0 (2024-05-16)
Ci
- ci: fixture in semantic release script (
5ce57e5)
Feature
- feat: support hot-reload for ui
If starting app with nrtk_explorer --hot-reload,
then a button appears in the upper right to reload
the ui modules and re-run Engine.ui method. (42ae90d)
Fix
- fix(nrtk_transforms): support new nrtk PybsmPerturber api
(cherry picked from commit efaf36a) (edea3f8)
- fix: change object_detector output
(cherry picked from commit 633e5f4) (fec5df0)
- fix(ImageDetection): fallback to Unknown name when uncatagorized
Horse drawn trolly thing has category ID of 0.
(cherry picked from commit f12cd7b) (75a236f)
- fix(object_detector): maintain input paths order of output predictions
Caused annotations to be overlaid on the wrong
images in image list view.
(cherry picked from commit fcc8f0a) (1b2fdb3)
- fix(benchmarks): allow using a external COCO ds
(cherry picked from commit 34b2fdb) (3e7e56e)
- fix(nrtk-explorer): multi platform paths
(cherry picked from commit 0131172) (cc631d7)
-
fix(nrtk_transforms): support new nrtk PybsmPerturber api (
efaf36a) -
fix: change object_detector output (
633e5f4) -
fix(ImageDetection): fallback to Unknown name when uncatagorized
Horse drawn trolly thing has category ID of 0. (f12cd7b)
- fix(object_detector): maintain input paths order of output predictions
Caused annotations to be overlaid on the wrong
images in image list view. (fcc8f0a)
-
fix(benchmarks): allow using a external COCO ds (
34b2fdb) -
fix(nrtk-explorer): multi platform paths (
0131172)
Refactor
- refactor(layout): factor out sections to functions (
dfa9a85)
Unknown
-
Merge main to release (
e32828d) -
Bump ubelt from 1.3.4 to 1.3.5
Bumps ubelt from 1.3.4 to 1.3.5.
updated-dependencies:
- dependency-name: ubelt
dependency-type: direct:production
update-type: version-update:semver-patch
...
Signed-off-by: dependabot[bot] <[email protected]> (87a292d)
- Bump scikit-learn from 1.4.1.post1 to 1.4.2
Bumps scikit-learn from 1.4.1.post1 to 1.4.2.
updated-dependencies:
- dependency-name: scikit-learn
dependency-type: direct:production
update-type: version-update:semver-patch
...
Signed-off-by: dependabot[bot] <[email protected]> (a8a4862)
- Auto-merge release back to main (
12ea17d)