Fixed uploading track annotations for multi-segment tasks #1396
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
azhavoro
changed the title
|
@azhavoro , do we have any related issues for the PR? |
nmanovic
reviewed
Apr 14, 2020
cvat/apps/annotation/annotation.py
Outdated
| splitted_data.tracks = [] | ||
|
|
||
| for track_ in self.tracks: | ||
| if not list(filter(is_frame_inside, track_['shapes'])): |
There was a problem hiding this comment.
I don't look how the function is used by for the slice function the condition looks wrong. Let's say we have a track with two shapes only. First shape is on 0 frame and second shape is on 100 frame (0 - start, and 100 - outside). Let's say the slice function is called with start=10 and stop=20. Looks like the track will be filtered out.
nmanovic
reviewed
Apr 14, 2020
nmanovic
reviewed
Apr 14, 2020
nmanovic
reviewed
Apr 14, 2020
nmanovic
reviewed
Apr 14, 2020
nmanovic
changed the title
nmanovic
reviewed
Apr 22, 2020
cvat/apps/engine/data_manager.py
Outdated
| @@ -290,6 +290,9 @@ def _modify_unmached_object(obj, end_frame): | |||
| shape["frame"] = end_frame | |||
| shape["outside"] = True | |||
| obj["shapes"].append(shape) | |||
| # Need to reset cached interpolated shapes | |||
| # because key shapes were changed | |||
| obj['interpolated_shapes'] = [] | |||
There was a problem hiding this comment.
@azhavoro , is it a performance issue from your point of view? Just want to understand if the solution is good or bad. In general there are two possible solutions here:
obj['interpolated_shapes'] = []- interpolate till the end (in this case it is trivial because just need to copy the shape till end_frame).
There was a problem hiding this comment.
Let's use the interpolation approach, updated.
nmanovic
reviewed
Apr 22, 2020
nmanovic
reviewed
Apr 22, 2020
nmanovic
reviewed
Apr 22, 2020
cvat/apps/annotation/annotation.py
Outdated
| @staticmethod | ||
| def _slice_track(track_, start, stop): | ||
| track = copy.deepcopy(track_) | ||
| segment_shapes = [s for s in track['shapes'] if AnnotationIR._is_shape_inside(s, start, stop)] |
There was a problem hiding this comment.
@azhavoro , if we have a track with only one shape with the frame less than start, segment_shapes will be empty.
There was a problem hiding this comment.
this case will be handled later in 113 line
There was a problem hiding this comment.
Potentially in the line we can get first shape with outside == True.
There was a problem hiding this comment.
added filter to remove starting shapes with outside==True after processing
nmanovic
requested changes
Apr 24, 2020
nmanovic
reviewed
Apr 27, 2020
nmanovic
reviewed
Apr 27, 2020
cvat/apps/annotation/annotation.py
Outdated
|
|
||
| for shape in interpolated_shapes: | ||
| if shape['frame'] == start and \ | ||
| (not segment_shapes or segment_shapes[0]['frame'] > start): |
There was a problem hiding this comment.
Is it possible that here shape will be with outside == True? Thus we will insert the shape but it looks wrong.
There was a problem hiding this comment.
added filter to remove starting shapes with outside==True after processing
nmanovic
reviewed
Apr 27, 2020
| @@ -77,13 +77,57 @@ def serialize(self): | |||
| if serializer.is_valid(raise_exception=True): | |||
| return serializer.data | |||
|
|
|||
| @staticmethod | |||
| def _is_shape_inside(shape, start, stop): | |||
| return start <= int(shape['frame']) <= stop | |||
There was a problem hiding this comment.
One tricky case here. Should we return True here for shapes with outside == True?
There was a problem hiding this comment.
added filter to remove starting shapes with outside==True after processing
Co-Authored-By: Nikita Manovich <[email protected]>
nmanovic
reviewed
Apr 27, 2020
| segment_shapes = segment_shapes[drop_shape_count:] | ||
| del track['interpolated_shapes'] | ||
| for shape in segment_shapes: | ||
| del shape['keyframe'] |
There was a problem hiding this comment.
Why should we delete the property for all segment_shapes? Could you please add a comment?
nmanovic
reviewed
Apr 27, 2020
cvat/apps/annotation/annotation.py
Outdated
| if len(segment_shapes) < len(track['shapes']): | ||
| interpolated_shapes = TrackManager.get_interpolated_shapes(track, start, stop) | ||
|
|
||
| for shape in interpolated_shapes: |
There was a problem hiding this comment.
@azhavoro , could you please think how to reorganize the code below? It looks messy.
It seems that get_interpolated_shapes doesn't respect start and stop by a reason. Thus let's filter them:
scoped_shapes = [s for s in interpolated_shapes['shapes'] if AnnotationIR._is_shape_inside(s, start, stop)]
if scoped_shapes and scoped_shapes[0]['outside']:
scoped_shapes.pop(0)
If scoped_shapes empty then nothing to do. Otherwise if not scoped_shapes[0]['keyframe'], let's add the shape into segment_shapes at the beginning. If not scoped_shapes[-1]['keyframe'], lets' add the shape into segment_shapes at the end.
Let's assign track['interpolated_shapes'] = scoped_shapes.
Of course I didn't check the logic. You need just improve your code and I hope the code above will give you some ideas how to do that.
azhavoro
changed the title
azhavoro
force-pushed
the
az/fix_uploading_annotation
branch
from
1b1ff78
to
f4ed055
Compare
azhavoro
changed the title
azhavoro
changed the title
azhavoro
changed the title
nmanovic
approved these changes
Apr 30, 2020
nmanovic
added a commit
that referenced
this pull request
May 29, 2020
* Added button to cancel started automatic annotation (#1198) * [WIP] Cuboid feature user guide (#1218) * Initial cuboid description * Added Gifs * Added gifs to descriptions * Formatting fixes * Codacy Fixes * Az/fix annotation dump upload (#1229) * fixed upload annotation in case of frame step != 1 * fixed upload annotation in case of attribute value is empty * Setup tag forward to the state * React UI: Added shortcuts (#1230) * [Datumaro] Label remapping transform (#1233) * Add label remapping transform * Apply transforms before project saving * Refactor voc converter * [Datumaro] Optimize mask operations (#1232) * Optimize mask to rle * Optimize mask operations * Fix dm format cmdline * Use RLE masks in datumaro format * Added tag support in new UI (without canvas drawing) * merge fix * Fixed copying/pasting actions * Deleted unused objects * [Datumaro] Dataset format auto detection (#1242) * Add dataset format detection * Add auto format detection for import * Split VOC extractor * Some debian package manager tweaks (#1235) * Some debian package manager tweaks By default, Ubuntu or Debian based "apt" or "apt-get" system installs recommended but not suggested packages . By passing "--no-install-recommends" option, the user lets apt-get know not to consider recommended packages as a dependency to install. This results in smaller downloads and installation of packages . Refer to blog at [Ubuntu Blog](https://ubuntu.com/blog/we-reduced-our-docker-images-by-60-with-no-install-recommends) . * doc: fix description of attribute CVAT XML format (#1168) * Fixed security issues in Datumaro (#1244) * Fixed security issues reported by bandit. * Fixed voc_format extractor * Sorted requirements, added a comment, removed nosec for exec. * Fix copying and creating tags * fixed git sync app (#1247) * fixed git sync app * removed shell=True for subprocess call * Fixed tags color changing and hiding * Fixed filters with tags * wip * PR fixed * Styles fixed * PR fixed * Update develop from release-0.6.0 branch (#1266) * temp * React UI: Attribute annotation mode (#1255) * Done main work * Fixed mount/unmount for canvas wrapper * Refactoring, added filters * Added missed file * Removed unnecessary useEffect * Removed extra code * Max 9 attributes, inputNumber -> Input in aam * Added blur * Renamed component * Fixed condition when validate number attribute * Some minor fixes * Fixed hotkeys config * Fixed canvas zoom * Improved behaviour of number & text * Fixed attributes switching order * Fix tags * Fixed interval * Installation issues for development environment (#1280) * Installation issues * Added ffmpeg * Bump acorn from 6.3.0 to 6.4.1 in /cvat-ui (#1270) * Bump acorn from 6.3.0 to 6.4.1 in /cvat-ui Bumps [acorn](https://github.com/acornjs/acorn) from 6.3.0 to 6.4.1. - [Release notes](https://github.com/acornjs/acorn/releases) - [Commits](acornjs/acorn@6.3.0...6.4.1) Signed-off-by: dependabot[bot] <[email protected]> * Updated CHANGELOG.md Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Boris Sekachev <[email protected]> * Bump acorn from 6.2.1 to 6.4.1 in /cvat-canvas (#1281) Bumps [acorn](https://github.com/acornjs/acorn) from 6.2.1 to 6.4.1. - [Release notes](https://github.com/acornjs/acorn/releases) - [Commits](acornjs/acorn@6.2.1...6.4.1) Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Use source label map for voc export (#1276) * Use source label map for voc export * Add line to changelog * [Datumaro] Fix frame matching in video annotations import (#1274) * Add extra frame matching way for videos * Add line to changelog * [Datumaro] Allow empty COCO dataset export (#1272) * Allow empty dataset export in coco * Add line to changelog Co-authored-by: Nikita Manovich <[email protected]> * [Datumaro] Fix occluded and z_order attributes export (#1271) * Fix occluded and z_order attributes export * Add line to changelog Co-authored-by: Nikita Manovich <[email protected]> * Fix LabelMe format (#1260) * Fix labelme filenames * Change module path * Add tests for LabelMe * Update test * Fix test * Add line in changelog * Added point deletion context menu * React UI: Added logging (#1288) * OpenVino 2020 (#1269) * added support for OpenVINO 2020 * fixed dextr and tf_annotation Co-authored-by: Andrey Zhavoronkov <[email protected]> * fixed point context menu for rectangles * Add recursive importers (#1290) * [Datumaro] MOT format (#1289) * Add mot format base * Add mot format * Extract common code * [Datumaro] LabelMe format (#1293) * Little refactoring * Add LabelMe format * [Datumaro] Update LabelMe format (#1296) * Little refactoring * Add LabelMe format * Add usernames * Update tests * Add extractor test * Add information about v0.6.1 release. * React UI: Better exception handling (#1297) * Fixed context menu on ubuntu * Fixed deleting of the latest point * fixes * Fix attributes with spaces in names (#1305) * fixed PR * [Datumaro] Fix image merging (#1301) * Always merge images for own dataset * Fix codacy * Validation for frame input value * Fixed UI fail when write characters in auto save interval input * Fixed input numbers in player settings * Fixed ui failing in propagate confirmation * Fixed latest input numbers, removed extra code, fixed typings * Fix navigation * Added undopoint in editing * Fixed: Could not receive frame (after merge on the latest frame) * Removed extra action dispatching * Which -> button property * Fixed: Inconsistent labels between UI and CLI/API * Fixed resize on right mouse button * Fixed create object URL after first save, fixed URL itself * Undo/redo returns frame where was a change (as it was done in previous version) * Fixed unit tests * [Datumaro] Extract common extractor functionality (#1319) * Extract common extractor functionality * Simplify coco extractor * Fix tfrecord * Fix AWS deployment (#1316) * Don't use antd less (big memory consumtion during the build process) * Fix AWS deployment guide * fix a problem with proxy and long domain names * remove sass loader for antd * Removed less and less-loader. * Simplified webpack config. * Data streaming using chunks (#1007) Huge feature (200+ commits from different developers). It completely changes layout of data (please expect very long DB migration process if you have a lot of tasks). The primary idea is to send data as zip chunks (e.g. 36 images in one chunk) or encoded video chunks and decode them on the client side. It helps to solve the problem with latency when you try to view a separate frame in the UI quickly (play mode). Another important feature of the patch is to provide access to the original images. Thus for annotations the client uses compressed chunks but if you want to export a dataset Datumaro will use original chunks (but video will be decoded with original quality and encoded with maximum/optimal quality in any case). * Shortcuts keymaps moved to state * Titles for objects in side menu * Fixed bug in menu * Titles in attribute annotations mode * Controls panel titles * Titles for object list header * Minor fixes * Added tooltips in top bar * Added settings tooltip * Optimized patch * Typos * Fix a problem with known hosts inside git app (cannot clone a repo from github.com) (#1330) * Fixed zOrder range computing in case when there are tags * Small preview and progress (#1331) * Reduce preview size (untested) * Fix tests * Improve media readers (untested) * fixed migration * fixed frame provider * fixed preview save * fixed stop frame * handle duration == None * codacy * added missed import * unified iteration over frames for media readers and fixed corner case when user specify stop_frame = 0 Co-authored-by: Nikita Manovich <[email protected]> * Move annotation formats to dataset manager (#1256) * Move formats to dataset manager * Unify datataset export and anno export implementations * Add track_id to TrackedShape, export tracked shapes * Replace MOT format * Replace LabelMe format * Add new formats to dm * Add dm tests * Extend TrackedShape * Enable dm test in CI * Fix tests * Add import * Fix tests * Fix mot track ids * Fix mot format * Update attribute logic in labelme tests * Use common code in yolo * Put datumaro in path in settings * Expect labels file in MOT next to annotations file * Add MOT format description * Add import * Add labelme format description * Linter fix * Linter fix2 * Compare attributes ordered * Update docs * Update tests * Az/fix migration (#1333) * fixed migration because readers interface changed * fixed tests * Fixed escape in draw * Insert multiple shapes * Added dialog window with some help info about filters * Typos in doc * Add missing information into changelog. * Add information for next release + update version. * Updated changelog * Updated changelog * Special behaviour for the attribute value __undefined__ * Fixed license year * No break space const * Updated changelog * Fixed year in license headers * Implementation of bitmap in client * Updated changelog * Z-layer support * Fixed settings after reopen a job * Do not show invisible objects on bitmap * Fix point interpolation (#1344) * Extend formats tests with different track types * Add unordered list comparison * Skip empty list comparison * fix * fix * Reproduce problem * Fix point interpolation for single point * undo rest api refactor * Added button to reset color settings * Updated changelog * Added option to display shape text always * Updated changelog * Hidden/outside fix * Fixed screen scaling * fixed dump error after moving format files (#1342) * fixed dump error after moving format files * updated changelog * Az/fix dextr (#1348) * Fixed dextr_segmentation app * updated changelog Co-authored-by: Nikita Manovich <[email protected]> * Disabled option by default * Fixed typos in interface * Increase preview size till 256, 256. Previous preview size was not optimal and led to a blurred image due to too small size. * Add a line into CHANGELOG.md about the change. * Refactor frame provider (#1355) * Refactor frame provider * fix * Add CODEOWNERS file (#1360) * Add CODEOWNERS file * Removed the outdated file. See https://github.com/opencv/cvat/graphs/contributors * Change codeowners * Add pull request and issue templates (#1359) * Add initial version of pull request template * Fix links * Fix codacy issues * Slightly improve titles of sections * Add a note about strikethough for the checklist. * Fix progress of a pull request (each checkbox is an issue) * Add the license header, checkboxes about the license. * Updated the license * Update the license to met https://github.com/licensee/licensee/blob/master/vendor/choosealicense.com/_licenses/mit.txt restrictions. * Fix the pull request template name * Make explaination text as comments (it will be visible when you edit the PR message) * Add initial version of the issue template. * Batch of fixes (#1370) * Some margins were change to paddings * Removed extra selected * Fix: added outside shapes when merge polyshapes * Fixed double scroll bars * Updated canvas table * Fixed setup methodf * Disabled change frame during drag, resize and editing * Fixed: hidden points are visible * Fixed: Merge is allowed for points, but clicks on points conflict with frame dragging logic * Fixed: do not filter removed objects * Updated CHANGELOG.md * Couple of headers updated * Added missed fields in exception logs (#1372) * Simplified codeowners file (#1376) * Fixed points visibility when go between frames * React UI: Added message when share is empty or not mounted (#1373) * Added message when share is empty or not mounted * Updated changelog * Update CHANGELOG.md Co-authored-by: Nikita Manovich <[email protected]> * Save full image paths in coco (#1381) * Add chunk iterator cache to frame provider (#1367) * Add chunk iterator cache * fix * Update item "Creating an annotation task" in User Guide (#1363) * Az/fix remote files (#1392) * fixed task creation from remote files * Update CHANGELOG.md * [Datumaro] Fix COCO keypoint export bug (#1388) * Instructions on using HTTPS (#1357) * Fix label comparison in voc format (#1382) * Update item "Interface of the annotation tool" in User Guide (#1386) * React UI: Batch of fixes (#1383) * Fixed: cannot read property 'set' of undefined * Fixed UI failing: save during drag/resize * Fixed multiple saving (shortcut sticking) * Undo/redo fixed * Allowed one interpolated point * Fixed API reaction when repository synchronization is failed * Updated changelog * Update item "Basic navigation" in User Guide and bug fix in user_guide.md (#1395) * Fix git app paths (#1400) * changed paths for the git repos * Update CHANGELOG.md * updated licence header * React UI: Displaying public ssh keys in UI (#1375) * Updated changelog * Typos * Batch of fixes (#1403) * Fixed bug when job cannot be opened * Fixed bug when deactivated shape is still highlighted * Fixed Error: 'AttributeError: 'tuple' object has no attribute 'read' * Fixed: wrong semi-automatic segmentation near edges of an image * Updated changelog Co-authored-by: Nikita Manovich <[email protected]> * React UI: Semi-automatic segmentation (#1398) * implemented checked * Implemented plugin * Added dialog windows * Updated changelo * Added cancel request * React UI: Automatic bordering for polygons and polylines during drawing/editing (#1394) * Fixed: cannot read property 'set' of undefined * Fixed UI failing: save during drag/resize * Fixed multiple saving (shortcut sticking) * Undo/redo fixed * Allowed one interpolated point * Fixed API reaction when repository synchronization is failed * Updated changelog * Auto bordering feature * Some fixes, added shortcuts * Fixed draw when start with one of supporting point * React UI: batch of fixes (#1404) * React UI has become a primary UI * Temporary disabled cuboid in cvat-core * Fixed: Failed to execute removChild.. (#1405) * Update CHANGELOG.md * Updated CHANGELOG (new version will be 1.0.0-beta.1) * Update CHANGELOG.md (next version is 1.0.0-beta.2) * Updated version of CVAT server till beta.2 * Fixed auto annotation, tf annotation and auto segmentation apps (#1409) * fixed code that uses FrameProvider, as the interface has changed * Update CHANGELOG.md * Update item "Types of shapes" (#1401) * React UI: ReID algorithm (#1406) * Initial commit * Connected storage * Added core API method * Done implementation * Removed rule * Removed double cancel * Updated changelog * Fixed: Cannot read property toFixed of undefined * Update CHANGELOG.md * Improve PR template (#1427) * Simplified PR template. * Remove a new line to make codacy happy. * fix: OSError:broken data stream (#1430) * [Datumaro] Fix duplicating keypoints in COCO export (#1435) * React UI: Fixed typos in remove annotations confirmation (#1450) * React UI: Batch of fixes (#1445) * Hide functionality (H) doesn't work * The highlighted attribute doesn't correspond to the chosen attribute in AAM * Inconvinient image shaking while drawing a polygon (hold Alt key during drawing/editing/grouping to drag an image) * Filter property "shape" doesn't work and extra operator in description * Block of text information doesn't disappear after deactivating for locked shapes * Annotation uploading fails in annotation view * UI freezes after canceling pasting with escape Co-authored-by: Nikita Manovich <[email protected]> * Update item "Annotation mode (basics)" in User Guide (#1412) * Update item "annotation mode (basics)" in User Guide * Replacing a gif017 with an image * React UI: Added client versioning (#1448) * Adjusted antd import * Wrapped core and canvas * Added versioning * Updated changelog, adjusted installation guide a bit * Update item "Interpolation mode (basics)" in User Guide (#1455) * Update item interpolation mode in user guide fix typos and contents * Fix a typo * Fix a typo in the filename Co-authored-by: Nikita Manovich <[email protected]> * React UI: batch of fixes (#1462) * CVAT new UI: add arrows on a mouse cursor * Delete point bug (in new UI) * fix auto annotation to not eat all RAM (#1328) Co-authored-by: Nikita Manovich <[email protected]> * Fixed versioning for node 8 (#1466) * Update items from AAM (basics) to Vocabulary in User Guide (#1459) * Update items from AAM (basics) to Vocabulary * Delete unused images and gif * Fix apache startup (#1467) * Fix Network Error after PC Restart (#1035) * Update CHANGELOG.md * Remove deprecated utils (#1477) * removed deprecated convert scripts * updated changelog * Fixed 'Open task' button doesn't work (#1474) * Fixed uploading track annotations for multi-segment tasks (#1396) * fixed uploading annotation for overlapped segments * fixed dump of tracks for multisegment task * Update CHANGELOG.md * fixed comments * fixed comments * Update cvat/apps/engine/data_manager.py Co-Authored-By: Nikita Manovich <[email protected]> * drop start shapes with outside==True for splitted track * code cleanup * fixed typo * fix Co-authored-by: Nikita Manovich <[email protected]> * React UI: cuboids (#1451) * Update changelog. * Update CHANGELOG and the version of CVAT server. * session.annotations.put() returns indexes of added objects (#1493) * session.annotations.put() returns indexes of added objects * Updated changelog * Updated README.md files (added info about versioning) (#1490) * Updated README.md files (added info about versioning) * Typos * Add coverage for python (#1483) * Fix coverage merging (#1504) * Updating instructions to serve Swagger documentation (#1502) * Update items Workspace and Types of shapes in User Guide (#1497) * Merge annotations and dataset_manager apps (#1352) * Fix coverage measurement (#1516) * fixed linter issues and store credentials cookie in the session object (#1526) * fixed issues * fixed issues and stored credentials cookies inside the session * fixed tests * React UI: batch of fixes (#1525) * Update item Settings in User Guide (#1508) * React UI: cookie policy drawer (#1511) * fixed linter issues (#1538) * fixed false tag activation (#1541) * Added item with npm package version increasing to the PR template (#1542) * [Datumaro] Fix coco import conflict with labels (#1548) * [Datumaro] Change alignment in mask parsing (#1547) * Include empty images in exported annotations (#1479) * Update item Bottom panel in User Guide (#1509) * Update item Bottom panel in User Guide rename the "Bottom panel" to the "Top panel" * Fixed typos, remove trailing spaces in user_guide.md * Fix user_guide.md and update image051.jpg * Ability to configure user agreements for the register user form (#1464) * Update item Side panel in User guide (#1513) * Update item Side panel in User guide * Removed trailing spaces in an Object sidebar item * Update cvat/apps/documentation/user_guide.md * fix typo * missing image correction Co-authored-by: Nikita Manovich <[email protected]> * Layout styles fixes * React UI: cuboid interpolation and cuboid drawing from rectangles (#1560) * Added backend cuboid interpolation and cuboid drawing from rectangles * Added CHANELOG.md * Fixed cuboid front edges stroke width * PR fixes * Fixed auto_segmentation app (#1562) * disabled tf eager execution for auto_segmentation * Update CHANGELOG.md * Add VOC grayscale masks test and documentation (#1576) * Add a test for unpainted masks * Update format documentation * [Datumaro] Fix mask to polygons warning (#1581) * Fix message, add test * update changelog * Fix cuboid conversion (#1577) * Fix cuboid conversion * update changelog Co-authored-by: Nikita Manovich <[email protected]> * [Datumaro] Simplify log level setting (#1583) * Simplify loglevel setting * update changelog * Fixed git synchronization (#1582) * fixed git synchronization * Update CHANGELOG.md Co-authored-by: Nikita Manovich <[email protected]> * Fixed dextr, fixed moving of the canvas (#1573) * Fixed dextr, fixed moving of the canvas * Updated CONTRIBUTUNG.md, updated version Co-authored-by: Nikita Manovich <[email protected]> * Added the ability to configure custom pageViewHit (may useful for web analytics) (#1566) * added the ability to configure custom pageViewHit (may useful for web analytics) * updated version and changelog * fixed comments * cvat-ui minor v++ * subscribe on history updates in the root component * updated Online Demo section of Readme (#1588) * updated Online Demo section of Readme * Change content of "online demo" section Co-authored-by: Nikita Manovich <[email protected]> * Fixed task creation for videos with uneven dimensions. (#1594) * used yuv420p format for compressed and original chunks * updated changelog * added settings to reduce access to analytics component (#1592) * added settings to reduce access to analytics component * updated CHANGELOG * fixed typo * Add item Controls sidebar in User Guide (#1510) * Update text, images and gif in user_guide.md (#1558) * Update user_guide.md, images and gif (#1556) Co-authored-by: Nikita Manovich <[email protected]> * update user_guide.md, images and gifs * Delete gif013 * add gif013 with correct name * Use z_order as a class property (#1589) * Use z_order as a class property * Fix z_order use in voc * Update changelog * Update item Annotation with polylines (#1596) * update user_guide.md and images * fix uppercase letters in images path in user_guide.md and remove trailing spaces * delete images and gifs containing uppercase letters in the name * add images with correct names * fix image paths in user_guide.md * Delete image133 * add image133 with correct name * Fix example yaml format (#1603) * delete duplicate item in user_guide.md (#1607) * update user_guide.md and add image (#1604) * update gifs, images and user_guide.md (#1605) * fix analytics permissions (#1608) * Update item Annotation with cuboids (#1598) * Update item Annotation with polygons in User guide v2 (#1612) * update user_guide.md * update images and gifs * Update Filter, Analytics and Shortcuts items in User Guide (#1606) * update user_guide.md and images * fix user_guide.md * delete unused image * delete unused images * Fix duplicate item in User Guide (#1617) * remove item Annotation with Auto Segmentation * fix link in user_guide.md * delete unused images * Slightly improve changelog * Update CVAT version Co-authored-by: Boris Sekachev <[email protected]> Co-authored-by: Tritin Truong <[email protected]> Co-authored-by: Andrey Zhavoronkov <[email protected]> Co-authored-by: Dmitry Kalinin <[email protected]> Co-authored-by: zhiltsov-max <[email protected]> Co-authored-by: Pratik Raj <[email protected]> Co-authored-by: Mathis Chenuet <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Boris Sekachev <[email protected]> Co-authored-by: Ben Hoff <[email protected]> Co-authored-by: Andrey Zhavoronkov <[email protected]> Co-authored-by: TOsmanov <[email protected]> Co-authored-by: ranko r sredojevic <[email protected]> Co-authored-by: Thomas Albrecht <[email protected]> Co-authored-by: Johannes222 <[email protected]> Co-authored-by: Gururaj Jeerge <[email protected]> Co-authored-by: timurx.osmanov <[email protected]> Co-authored-by: YutaYamazaki <[email protected]>
frndmg
pushed a commit
to signatrix/cvat
that referenced
this pull request
Aug 5, 2020
* Added button to cancel started automatic annotation (cvat-ai#1198) * [WIP] Cuboid feature user guide (cvat-ai#1218) * Initial cuboid description * Added Gifs * Added gifs to descriptions * Formatting fixes * Codacy Fixes * Az/fix annotation dump upload (cvat-ai#1229) * fixed upload annotation in case of frame step != 1 * fixed upload annotation in case of attribute value is empty * Setup tag forward to the state * React UI: Added shortcuts (cvat-ai#1230) * [Datumaro] Label remapping transform (cvat-ai#1233) * Add label remapping transform * Apply transforms before project saving * Refactor voc converter * [Datumaro] Optimize mask operations (cvat-ai#1232) * Optimize mask to rle * Optimize mask operations * Fix dm format cmdline * Use RLE masks in datumaro format * Added tag support in new UI (without canvas drawing) * merge fix * Fixed copying/pasting actions * Deleted unused objects * [Datumaro] Dataset format auto detection (cvat-ai#1242) * Add dataset format detection * Add auto format detection for import * Split VOC extractor * Some debian package manager tweaks (cvat-ai#1235) * Some debian package manager tweaks By default, Ubuntu or Debian based "apt" or "apt-get" system installs recommended but not suggested packages . By passing "--no-install-recommends" option, the user lets apt-get know not to consider recommended packages as a dependency to install. This results in smaller downloads and installation of packages . Refer to blog at [Ubuntu Blog](https://ubuntu.com/blog/we-reduced-our-docker-images-by-60-with-no-install-recommends) . * doc: fix description of attribute CVAT XML format (cvat-ai#1168) * Fixed security issues in Datumaro (cvat-ai#1244) * Fixed security issues reported by bandit. * Fixed voc_format extractor * Sorted requirements, added a comment, removed nosec for exec. * Fix copying and creating tags * fixed git sync app (cvat-ai#1247) * fixed git sync app * removed shell=True for subprocess call * Fixed tags color changing and hiding * Fixed filters with tags * wip * PR fixed * Styles fixed * PR fixed * Update develop from release-0.6.0 branch (cvat-ai#1266) * temp * React UI: Attribute annotation mode (cvat-ai#1255) * Done main work * Fixed mount/unmount for canvas wrapper * Refactoring, added filters * Added missed file * Removed unnecessary useEffect * Removed extra code * Max 9 attributes, inputNumber -> Input in aam * Added blur * Renamed component * Fixed condition when validate number attribute * Some minor fixes * Fixed hotkeys config * Fixed canvas zoom * Improved behaviour of number & text * Fixed attributes switching order * Fix tags * Fixed interval * Installation issues for development environment (cvat-ai#1280) * Installation issues * Added ffmpeg * Bump acorn from 6.3.0 to 6.4.1 in /cvat-ui (cvat-ai#1270) * Bump acorn from 6.3.0 to 6.4.1 in /cvat-ui Bumps [acorn](https://github.com/acornjs/acorn) from 6.3.0 to 6.4.1. - [Release notes](https://github.com/acornjs/acorn/releases) - [Commits](acornjs/acorn@6.3.0...6.4.1) Signed-off-by: dependabot[bot] <[email protected]> * Updated CHANGELOG.md Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Boris Sekachev <[email protected]> * Bump acorn from 6.2.1 to 6.4.1 in /cvat-canvas (cvat-ai#1281) Bumps [acorn](https://github.com/acornjs/acorn) from 6.2.1 to 6.4.1. - [Release notes](https://github.com/acornjs/acorn/releases) - [Commits](acornjs/acorn@6.2.1...6.4.1) Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Use source label map for voc export (cvat-ai#1276) * Use source label map for voc export * Add line to changelog * [Datumaro] Fix frame matching in video annotations import (cvat-ai#1274) * Add extra frame matching way for videos * Add line to changelog * [Datumaro] Allow empty COCO dataset export (cvat-ai#1272) * Allow empty dataset export in coco * Add line to changelog Co-authored-by: Nikita Manovich <[email protected]> * [Datumaro] Fix occluded and z_order attributes export (cvat-ai#1271) * Fix occluded and z_order attributes export * Add line to changelog Co-authored-by: Nikita Manovich <[email protected]> * Fix LabelMe format (cvat-ai#1260) * Fix labelme filenames * Change module path * Add tests for LabelMe * Update test * Fix test * Add line in changelog * Added point deletion context menu * React UI: Added logging (cvat-ai#1288) * OpenVino 2020 (cvat-ai#1269) * added support for OpenVINO 2020 * fixed dextr and tf_annotation Co-authored-by: Andrey Zhavoronkov <[email protected]> * fixed point context menu for rectangles * Add recursive importers (cvat-ai#1290) * [Datumaro] MOT format (cvat-ai#1289) * Add mot format base * Add mot format * Extract common code * [Datumaro] LabelMe format (cvat-ai#1293) * Little refactoring * Add LabelMe format * [Datumaro] Update LabelMe format (cvat-ai#1296) * Little refactoring * Add LabelMe format * Add usernames * Update tests * Add extractor test * Add information about v0.6.1 release. * React UI: Better exception handling (cvat-ai#1297) * Fixed context menu on ubuntu * Fixed deleting of the latest point * fixes * Fix attributes with spaces in names (cvat-ai#1305) * fixed PR * [Datumaro] Fix image merging (cvat-ai#1301) * Always merge images for own dataset * Fix codacy * Validation for frame input value * Fixed UI fail when write characters in auto save interval input * Fixed input numbers in player settings * Fixed ui failing in propagate confirmation * Fixed latest input numbers, removed extra code, fixed typings * Fix navigation * Added undopoint in editing * Fixed: Could not receive frame (after merge on the latest frame) * Removed extra action dispatching * Which -> button property * Fixed: Inconsistent labels between UI and CLI/API * Fixed resize on right mouse button * Fixed create object URL after first save, fixed URL itself * Undo/redo returns frame where was a change (as it was done in previous version) * Fixed unit tests * [Datumaro] Extract common extractor functionality (cvat-ai#1319) * Extract common extractor functionality * Simplify coco extractor * Fix tfrecord * Fix AWS deployment (cvat-ai#1316) * Don't use antd less (big memory consumtion during the build process) * Fix AWS deployment guide * fix a problem with proxy and long domain names * remove sass loader for antd * Removed less and less-loader. * Simplified webpack config. * Data streaming using chunks (cvat-ai#1007) Huge feature (200+ commits from different developers). It completely changes layout of data (please expect very long DB migration process if you have a lot of tasks). The primary idea is to send data as zip chunks (e.g. 36 images in one chunk) or encoded video chunks and decode them on the client side. It helps to solve the problem with latency when you try to view a separate frame in the UI quickly (play mode). Another important feature of the patch is to provide access to the original images. Thus for annotations the client uses compressed chunks but if you want to export a dataset Datumaro will use original chunks (but video will be decoded with original quality and encoded with maximum/optimal quality in any case). * Shortcuts keymaps moved to state * Titles for objects in side menu * Fixed bug in menu * Titles in attribute annotations mode * Controls panel titles * Titles for object list header * Minor fixes * Added tooltips in top bar * Added settings tooltip * Optimized patch * Typos * Fix a problem with known hosts inside git app (cannot clone a repo from github.com) (cvat-ai#1330) * Fixed zOrder range computing in case when there are tags * Small preview and progress (cvat-ai#1331) * Reduce preview size (untested) * Fix tests * Improve media readers (untested) * fixed migration * fixed frame provider * fixed preview save * fixed stop frame * handle duration == None * codacy * added missed import * unified iteration over frames for media readers and fixed corner case when user specify stop_frame = 0 Co-authored-by: Nikita Manovich <[email protected]> * Move annotation formats to dataset manager (cvat-ai#1256) * Move formats to dataset manager * Unify datataset export and anno export implementations * Add track_id to TrackedShape, export tracked shapes * Replace MOT format * Replace LabelMe format * Add new formats to dm * Add dm tests * Extend TrackedShape * Enable dm test in CI * Fix tests * Add import * Fix tests * Fix mot track ids * Fix mot format * Update attribute logic in labelme tests * Use common code in yolo * Put datumaro in path in settings * Expect labels file in MOT next to annotations file * Add MOT format description * Add import * Add labelme format description * Linter fix * Linter fix2 * Compare attributes ordered * Update docs * Update tests * Az/fix migration (cvat-ai#1333) * fixed migration because readers interface changed * fixed tests * Fixed escape in draw * Insert multiple shapes * Added dialog window with some help info about filters * Typos in doc * Add missing information into changelog. * Add information for next release + update version. * Updated changelog * Updated changelog * Special behaviour for the attribute value __undefined__ * Fixed license year * No break space const * Updated changelog * Fixed year in license headers * Implementation of bitmap in client * Updated changelog * Z-layer support * Fixed settings after reopen a job * Do not show invisible objects on bitmap * Fix point interpolation (cvat-ai#1344) * Extend formats tests with different track types * Add unordered list comparison * Skip empty list comparison * fix * fix * Reproduce problem * Fix point interpolation for single point * undo rest api refactor * Added button to reset color settings * Updated changelog * Added option to display shape text always * Updated changelog * Hidden/outside fix * Fixed screen scaling * fixed dump error after moving format files (cvat-ai#1342) * fixed dump error after moving format files * updated changelog * Az/fix dextr (cvat-ai#1348) * Fixed dextr_segmentation app * updated changelog Co-authored-by: Nikita Manovich <[email protected]> * Disabled option by default * Fixed typos in interface * Increase preview size till 256, 256. Previous preview size was not optimal and led to a blurred image due to too small size. * Add a line into CHANGELOG.md about the change. * Refactor frame provider (cvat-ai#1355) * Refactor frame provider * fix * Add CODEOWNERS file (cvat-ai#1360) * Add CODEOWNERS file * Removed the outdated file. See https://github.com/opencv/cvat/graphs/contributors * Change codeowners * Add pull request and issue templates (cvat-ai#1359) * Add initial version of pull request template * Fix links * Fix codacy issues * Slightly improve titles of sections * Add a note about strikethough for the checklist. * Fix progress of a pull request (each checkbox is an issue) * Add the license header, checkboxes about the license. * Updated the license * Update the license to met https://github.com/licensee/licensee/blob/master/vendor/choosealicense.com/_licenses/mit.txt restrictions. * Fix the pull request template name * Make explaination text as comments (it will be visible when you edit the PR message) * Add initial version of the issue template. * Batch of fixes (cvat-ai#1370) * Some margins were change to paddings * Removed extra selected * Fix: added outside shapes when merge polyshapes * Fixed double scroll bars * Updated canvas table * Fixed setup methodf * Disabled change frame during drag, resize and editing * Fixed: hidden points are visible * Fixed: Merge is allowed for points, but clicks on points conflict with frame dragging logic * Fixed: do not filter removed objects * Updated CHANGELOG.md * Couple of headers updated * Added missed fields in exception logs (cvat-ai#1372) * Simplified codeowners file (cvat-ai#1376) * Fixed points visibility when go between frames * React UI: Added message when share is empty or not mounted (cvat-ai#1373) * Added message when share is empty or not mounted * Updated changelog * Update CHANGELOG.md Co-authored-by: Nikita Manovich <[email protected]> * Save full image paths in coco (cvat-ai#1381) * Add chunk iterator cache to frame provider (cvat-ai#1367) * Add chunk iterator cache * fix * Update item "Creating an annotation task" in User Guide (cvat-ai#1363) * Az/fix remote files (cvat-ai#1392) * fixed task creation from remote files * Update CHANGELOG.md * [Datumaro] Fix COCO keypoint export bug (cvat-ai#1388) * Instructions on using HTTPS (cvat-ai#1357) * Fix label comparison in voc format (cvat-ai#1382) * Update item "Interface of the annotation tool" in User Guide (cvat-ai#1386) * React UI: Batch of fixes (cvat-ai#1383) * Fixed: cannot read property 'set' of undefined * Fixed UI failing: save during drag/resize * Fixed multiple saving (shortcut sticking) * Undo/redo fixed * Allowed one interpolated point * Fixed API reaction when repository synchronization is failed * Updated changelog * Update item "Basic navigation" in User Guide and bug fix in user_guide.md (cvat-ai#1395) * Fix git app paths (cvat-ai#1400) * changed paths for the git repos * Update CHANGELOG.md * updated licence header * React UI: Displaying public ssh keys in UI (cvat-ai#1375) * Updated changelog * Typos * Batch of fixes (cvat-ai#1403) * Fixed bug when job cannot be opened * Fixed bug when deactivated shape is still highlighted * Fixed Error: 'AttributeError: 'tuple' object has no attribute 'read' * Fixed: wrong semi-automatic segmentation near edges of an image * Updated changelog Co-authored-by: Nikita Manovich <[email protected]> * React UI: Semi-automatic segmentation (cvat-ai#1398) * implemented checked * Implemented plugin * Added dialog windows * Updated changelo * Added cancel request * React UI: Automatic bordering for polygons and polylines during drawing/editing (cvat-ai#1394) * Fixed: cannot read property 'set' of undefined * Fixed UI failing: save during drag/resize * Fixed multiple saving (shortcut sticking) * Undo/redo fixed * Allowed one interpolated point * Fixed API reaction when repository synchronization is failed * Updated changelog * Auto bordering feature * Some fixes, added shortcuts * Fixed draw when start with one of supporting point * React UI: batch of fixes (cvat-ai#1404) * React UI has become a primary UI * Temporary disabled cuboid in cvat-core * Fixed: Failed to execute removChild.. (cvat-ai#1405) * Update CHANGELOG.md * Updated CHANGELOG (new version will be 1.0.0-beta.1) * Update CHANGELOG.md (next version is 1.0.0-beta.2) * Updated version of CVAT server till beta.2 * Fixed auto annotation, tf annotation and auto segmentation apps (cvat-ai#1409) * fixed code that uses FrameProvider, as the interface has changed * Update CHANGELOG.md * Update item "Types of shapes" (cvat-ai#1401) * React UI: ReID algorithm (cvat-ai#1406) * Initial commit * Connected storage * Added core API method * Done implementation * Removed rule * Removed double cancel * Updated changelog * Fixed: Cannot read property toFixed of undefined * Update CHANGELOG.md * Improve PR template (cvat-ai#1427) * Simplified PR template. * Remove a new line to make codacy happy. * fix: OSError:broken data stream (cvat-ai#1430) * [Datumaro] Fix duplicating keypoints in COCO export (cvat-ai#1435) * React UI: Fixed typos in remove annotations confirmation (cvat-ai#1450) * React UI: Batch of fixes (cvat-ai#1445) * Hide functionality (H) doesn't work * The highlighted attribute doesn't correspond to the chosen attribute in AAM * Inconvinient image shaking while drawing a polygon (hold Alt key during drawing/editing/grouping to drag an image) * Filter property "shape" doesn't work and extra operator in description * Block of text information doesn't disappear after deactivating for locked shapes * Annotation uploading fails in annotation view * UI freezes after canceling pasting with escape Co-authored-by: Nikita Manovich <[email protected]> * Update item "Annotation mode (basics)" in User Guide (cvat-ai#1412) * Update item "annotation mode (basics)" in User Guide * Replacing a gif017 with an image * React UI: Added client versioning (cvat-ai#1448) * Adjusted antd import * Wrapped core and canvas * Added versioning * Updated changelog, adjusted installation guide a bit * Update item "Interpolation mode (basics)" in User Guide (cvat-ai#1455) * Update item interpolation mode in user guide fix typos and contents * Fix a typo * Fix a typo in the filename Co-authored-by: Nikita Manovich <[email protected]> * React UI: batch of fixes (cvat-ai#1462) * CVAT new UI: add arrows on a mouse cursor * Delete point bug (in new UI) * fix auto annotation to not eat all RAM (cvat-ai#1328) Co-authored-by: Nikita Manovich <[email protected]> * Fixed versioning for node 8 (cvat-ai#1466) * Update items from AAM (basics) to Vocabulary in User Guide (cvat-ai#1459) * Update items from AAM (basics) to Vocabulary * Delete unused images and gif * Fix apache startup (cvat-ai#1467) * Fix Network Error after PC Restart (cvat-ai#1035) * Update CHANGELOG.md * Remove deprecated utils (cvat-ai#1477) * removed deprecated convert scripts * updated changelog * Fixed 'Open task' button doesn't work (cvat-ai#1474) * Fixed uploading track annotations for multi-segment tasks (cvat-ai#1396) * fixed uploading annotation for overlapped segments * fixed dump of tracks for multisegment task * Update CHANGELOG.md * fixed comments * fixed comments * Update cvat/apps/engine/data_manager.py Co-Authored-By: Nikita Manovich <[email protected]> * drop start shapes with outside==True for splitted track * code cleanup * fixed typo * fix Co-authored-by: Nikita Manovich <[email protected]> * React UI: cuboids (cvat-ai#1451) * Update changelog. * Update CHANGELOG and the version of CVAT server. * session.annotations.put() returns indexes of added objects (cvat-ai#1493) * session.annotations.put() returns indexes of added objects * Updated changelog * Updated README.md files (added info about versioning) (cvat-ai#1490) * Updated README.md files (added info about versioning) * Typos * Add coverage for python (cvat-ai#1483) * Fix coverage merging (cvat-ai#1504) * Updating instructions to serve Swagger documentation (cvat-ai#1502) * Update items Workspace and Types of shapes in User Guide (cvat-ai#1497) * Merge annotations and dataset_manager apps (cvat-ai#1352) * Fix coverage measurement (cvat-ai#1516) * fixed linter issues and store credentials cookie in the session object (cvat-ai#1526) * fixed issues * fixed issues and stored credentials cookies inside the session * fixed tests * React UI: batch of fixes (cvat-ai#1525) * Update item Settings in User Guide (cvat-ai#1508) * React UI: cookie policy drawer (cvat-ai#1511) * fixed linter issues (cvat-ai#1538) * fixed false tag activation (cvat-ai#1541) * Added item with npm package version increasing to the PR template (cvat-ai#1542) * [Datumaro] Fix coco import conflict with labels (cvat-ai#1548) * [Datumaro] Change alignment in mask parsing (cvat-ai#1547) * Include empty images in exported annotations (cvat-ai#1479) * Update item Bottom panel in User Guide (cvat-ai#1509) * Update item Bottom panel in User Guide rename the "Bottom panel" to the "Top panel" * Fixed typos, remove trailing spaces in user_guide.md * Fix user_guide.md and update image051.jpg * Ability to configure user agreements for the register user form (cvat-ai#1464) * Update item Side panel in User guide (cvat-ai#1513) * Update item Side panel in User guide * Removed trailing spaces in an Object sidebar item * Update cvat/apps/documentation/user_guide.md * fix typo * missing image correction Co-authored-by: Nikita Manovich <[email protected]> * Layout styles fixes * React UI: cuboid interpolation and cuboid drawing from rectangles (cvat-ai#1560) * Added backend cuboid interpolation and cuboid drawing from rectangles * Added CHANELOG.md * Fixed cuboid front edges stroke width * PR fixes * Fixed auto_segmentation app (cvat-ai#1562) * disabled tf eager execution for auto_segmentation * Update CHANGELOG.md * Add VOC grayscale masks test and documentation (cvat-ai#1576) * Add a test for unpainted masks * Update format documentation * [Datumaro] Fix mask to polygons warning (cvat-ai#1581) * Fix message, add test * update changelog * Fix cuboid conversion (cvat-ai#1577) * Fix cuboid conversion * update changelog Co-authored-by: Nikita Manovich <[email protected]> * [Datumaro] Simplify log level setting (cvat-ai#1583) * Simplify loglevel setting * update changelog * Fixed git synchronization (cvat-ai#1582) * fixed git synchronization * Update CHANGELOG.md Co-authored-by: Nikita Manovich <[email protected]> * Fixed dextr, fixed moving of the canvas (cvat-ai#1573) * Fixed dextr, fixed moving of the canvas * Updated CONTRIBUTUNG.md, updated version Co-authored-by: Nikita Manovich <[email protected]> * Added the ability to configure custom pageViewHit (may useful for web analytics) (cvat-ai#1566) * added the ability to configure custom pageViewHit (may useful for web analytics) * updated version and changelog * fixed comments * cvat-ui minor v++ * subscribe on history updates in the root component * updated Online Demo section of Readme (cvat-ai#1588) * updated Online Demo section of Readme * Change content of "online demo" section Co-authored-by: Nikita Manovich <[email protected]> * Fixed task creation for videos with uneven dimensions. (cvat-ai#1594) * used yuv420p format for compressed and original chunks * updated changelog * added settings to reduce access to analytics component (cvat-ai#1592) * added settings to reduce access to analytics component * updated CHANGELOG * fixed typo * Add item Controls sidebar in User Guide (cvat-ai#1510) * Update text, images and gif in user_guide.md (cvat-ai#1558) * Update user_guide.md, images and gif (cvat-ai#1556) Co-authored-by: Nikita Manovich <[email protected]> * update user_guide.md, images and gifs * Delete gif013 * add gif013 with correct name * Use z_order as a class property (cvat-ai#1589) * Use z_order as a class property * Fix z_order use in voc * Update changelog * Update item Annotation with polylines (cvat-ai#1596) * update user_guide.md and images * fix uppercase letters in images path in user_guide.md and remove trailing spaces * delete images and gifs containing uppercase letters in the name * add images with correct names * fix image paths in user_guide.md * Delete image133 * add image133 with correct name * Fix example yaml format (cvat-ai#1603) * delete duplicate item in user_guide.md (cvat-ai#1607) * update user_guide.md and add image (cvat-ai#1604) * update gifs, images and user_guide.md (cvat-ai#1605) * fix analytics permissions (cvat-ai#1608) * Update item Annotation with cuboids (cvat-ai#1598) * Update item Annotation with polygons in User guide v2 (cvat-ai#1612) * update user_guide.md * update images and gifs * Update Filter, Analytics and Shortcuts items in User Guide (cvat-ai#1606) * update user_guide.md and images * fix user_guide.md * delete unused image * delete unused images * Fix duplicate item in User Guide (cvat-ai#1617) * remove item Annotation with Auto Segmentation * fix link in user_guide.md * delete unused images * Slightly improve changelog * Update CVAT version Co-authored-by: Boris Sekachev <[email protected]> Co-authored-by: Tritin Truong <[email protected]> Co-authored-by: Andrey Zhavoronkov <[email protected]> Co-authored-by: Dmitry Kalinin <[email protected]> Co-authored-by: zhiltsov-max <[email protected]> Co-authored-by: Pratik Raj <[email protected]> Co-authored-by: Mathis Chenuet <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Boris Sekachev <[email protected]> Co-authored-by: Ben Hoff <[email protected]> Co-authored-by: Andrey Zhavoronkov <[email protected]> Co-authored-by: TOsmanov <[email protected]> Co-authored-by: ranko r sredojevic <[email protected]> Co-authored-by: Thomas Albrecht <[email protected]> Co-authored-by: Johannes222 <[email protected]> Co-authored-by: Gururaj Jeerge <[email protected]> Co-authored-by: timurx.osmanov <[email protected]> Co-authored-by: YutaYamazaki <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixed uploading track annotations for multi-segment tasks
resolve #1313
Motivation and context
Uploading track annotation in case of multi-segment task works incorrectly because imported tracked shapes is not filtered by frame and it leads to artifacts in the old UI and wrong dump file.
Also fixed bug with dump when track in the dump file may have no
closingshape withoutside=True.How has this been tested?
tested manually
Checklist
License
Feel free to contact the maintainers if that's a concern.