cvat

Fixed

• Formats with the custom track_id attribute should import outside track shapes properly (e.g. COCO, COCO Keypoints, Datumaro, PASCAL VOC) (https://github.com/opencv/cvat/pull/7669)

• Inefficient resources fetching in admin panel leading to 504 Gateway Timeout (https://github.com/cvat-ai/cvat/pull/7767)

• Optimized memory usage when retrieving annotations by disabling internal Django QuerySet caching (https://github.com/cvat-ai/cvat/pull/7748)

• Annotations are not shown on the 0 frame sometimes (https://github.com/cvat-ai/cvat/pull/7796)

• Extra requests in PolicyEnforcer when at least one policy is rejected, others are not checked (https://github.com/cvat-ai/cvat/pull/7803)

• Project's updated_date was not updated after changing annotations in jobs (https://github.com/cvat-ai/cvat/pull/7808)

Added

• Number of objects on the frame is shown on the right sidebar (https://github.com/opencv/cvat/pull/7654)

• Shortcut to switch "pinned" property (P) (https://github.com/opencv/cvat/pull/7709)

• Support for .rar, .tar, .gz, .bz2, .cpio, .7z archives (https://github.com/opencv/cvat/pull/7729)

Changed

• Updated links to the documentation website to point to the new domain, docs.cvat.ai (https://github.com/cvat-ai/cvat/pull/7722)

• Job and task download_frames now accepts custom extension for images (https://github.com/cvat-ai/cvat/pull/7697)

Fixed

• Creating tasks with special characters in uploaded filename (https://github.com/opencv/cvat/pull/7646)

• Find next frame with issues ignored hide resolved issues setting (https://github.com/opencv/cvat/pull/7653)

• Objects menu is invisible for GT objects in GT job (https://github.com/opencv/cvat/pull/7714)

• Missing RegisterSerializerEx email_verification_required and key parameters now are included in the server schema (https://github.com/cvat-ai/cvat/pull/7635)

• Standardize the alignment of empty-list components (https://github.com/opencv/cvat/pull/7659)

• Labels in WiderFace dataset example (https://github.com/opencv/cvat/pull/7716)

• Export without images in Datumaro format - no empty "media" and "point_cloud" fields should be present (https://github.com/opencv/cvat/pull/7716)

• Fixed the inability to rename label attributes after creating them. (https://github.com/cvat-ai/cvat/pull/7670)

• When user starts editing a mask, it becomes smoother (not pixelated) (https://github.com/cvat-ai/cvat/pull/7747)

Added

• Tooltips for long names on cards (projects, tasks, cloud storages, and models) (https://github.com/opencv/cvat/pull/7550)

Removed

• The POST /api/tasks/{id}/data endpoint no longer accepts several parameters that didn't have any useful function: size, compressed_chunk_type, original_chunk_type (https://github.com/opencv/cvat/pull/7663)

Fixed

• Duplicated notifications for automatic annotation (https://github.com/opencv/cvat/pull/7595)

• Made quality report update job scheduling more efficient (https://github.com/opencv/cvat/pull/7596)

• Incorrect file name usage when importing annotations from a cloud storage (https://github.com/opencv/cvat/pull/7599)

• Using single shape annotation mode with multiple labels (https://github.com/opencv/cvat/pull/7606)

• Part of sidebar not visible in attribute annotation mode when there are a lot of attribute values (https://github.com/opencv/cvat/pull/7610)

• Changed interpolation behavior in annotation.py, now correctly keep the last frame

• Insert last frame if it is key to the track, fixes data corruption when tracks crossing more than 1 jobs (https://github.com/opencv/cvat/pull/7615)

• Label constructor validation of empty label names (https://github.com/opencv/cvat/pull/7627)

• Incorrect alignment of empty job list component (https://github.com/opencv/cvat/pull/7621)

• Remove underlying pixels feature is not applied immediately (https://github.com/opencv/cvat/pull/7637)

• Corrected the formula for per-class accuracy in quality reports; the old formula is now exposed as the jaccard_index key (https://github.com/opencv/cvat/pull/7640)

• Sending /events request from logged-out user (https://github.com/opencv/cvat/pull/7608)

• Fixed accuracy being displayed incorrectly on the task analytics page (https://github.com/opencv/cvat/pull/7652)

• Fixed an invalid default overlap size being selected for video tasks with small segments (https://github.com/opencv/cvat/pull/7681)

• Fixed redundant jobs being created for tasks with non-zero overlap in certain cases (https://github.com/opencv/cvat/pull/7681)

• Accumulation of confusion matrix across all jobs in a task when creating a quality report (https://github.com/opencv/cvat/pull/7604)

• 90 deg-rotated video was added with "Prefer Zip Chunks" disabled was warped, fixed using the static cropImage function. (https://github.com/opencv/cvat/pull/7583)

Changed

• Sped up resource updates when there are no matching webhooks (https://github.com/opencv/cvat/pull/7553)

Fixed

• Job and task updated_date are no longer bumped twice when updating annotations (https://github.com/opencv/cvat/pull/7556)

• Sending PATCH /jobs/{id}/data/meta on each job save even if nothing changed in meta data (https://github.com/opencv/cvat/pull/7560)

• Sending GET /jobs/{id}/data/meta twice on each job load (https://github.com/opencv/cvat/pull/7560)

• Made analytics report update job scheduling more efficient (https://github.com/opencv/cvat/pull/7576)

• Fixed being unable to connect to in-mem Redis when the password includes URL-unsafe characters (https://github.com/opencv/cvat/pull/7577)

• Segment anything decoder is loaded anytime when CVAT is opened, but might be not required (https://github.com/opencv/cvat/pull/7564)

Added

• Single shape annotation mode allowing to easily annotate scenarious where a user only needs to draw one object on one image (https://github.com/opencv/cvat/pull/7486)

Fixed

• Fixed a problem with Korean/Chinese characters in attribute annotation mode (https://github.com/opencv/cvat/pull/7380)

• Fixed incorrect working time calculation in the case where an event occurred during another event (https://github.com/opencv/cvat/pull/7511)

• Fixed working time not being calculated for the first event in each batch sent from the UI (https://github.com/opencv/cvat/pull/7511)

• Submit button is enabled while creating a ground truth job (https://github.com/opencv/cvat/pull/7540)

Added

• Added dataset:export and dataset:import events that are logged when the user initiates an export or import of a project, task or job (https://github.com/opencv/cvat/pull/7476)

Changed

• Now menus in the web interface are triggered by click, not by hover as before (https://github.com/opencv/cvat/pull/7431)

Removed

• Removed support for the TFRecord dataset format (https://github.com/opencv/cvat/pull/7416)

Fixed

• On quality page for a task, only the first page with jobs has quality report metrics (https://github.com/opencv/cvat/pull/7441)

• Side effects of data changes, such as the sending of webhooks, are no longer triggered until after the changes have been committed to the database (https://github.com/opencv/cvat/pull/7460, https://github.com/opencv/cvat/pull/7477)

Changed

• The "message" field of the payload of send:exception events no longer includes a trailing linebreak (https://github.com/opencv/cvat/pull/7407)

• Annotation guide is opened automatically if not seen yet when the job is "new annotation" (https://github.com/opencv/cvat/pull/7410)

• Annotation guide will be opened automatically if this is specified in a link /tasks/<id>/jobs/<id>?openGuide (https://github.com/opencv/cvat/pull/7410)

• Reduced number of server requests, made by clients (https://github.com/opencv/cvat/pull/7446)

• Server exception rest_framework.exceptions.NotAuthenticated is not logged by analytics anymore (https://github.com/opencv/cvat/pull/7457)

Fixed

• Prevented zombie processes from accumulating in the Kvrocks container (https://github.com/opencv/cvat/pull/7412)

• Fix Redis exceptions crashing the /api/server/health/ endpoint (https://github.com/opencv/cvat/pull/7417)

• Unhandled exception "Cannot read properties of null (reading 'plot')" (https://github.com/opencv/cvat/pull/7422)

• Unhandled exception "Cannot read properties of undefined (reading 'toLowerCase')" (https://github.com/opencv/cvat/pull/7421)

Changed

• Enhanced errors messaging for better perception by users (https://github.com/opencv/cvat/pull/7331)

Fixed

• Empty masks might be created with polygon-minus tool (https://github.com/opencv/cvat/pull/7295)

• Empty masks might be created as a result of removing underlying pixels (https://github.com/opencv/cvat/pull/7295)

• Fixed excessive memory usage when exporting a project with multiple video tasks (https://github.com/opencv/cvat/pull/7374)

• OpenCV tracker MIL works one frame behind (https://github.com/opencv/cvat/pull/7399)

Changed

• KeyDB used as data cache replaced by Kvrocks (https://github.com/opencv/cvat/pull/7339)

Fixed

• 504 Timeout error when exporting resources to cloud storage (https://github.com/opencv/cvat/pull/7317)

• Enqueuing deferred jobs when their dependencies have been started -> cancelled -> restarted -> finished (https://github.com/opencv/cvat/pull/7317)

• UI failed when open context menu for a skeleton element on a frame with a conflict (https://github.com/opencv/cvat/pull/7362)

• Issue can not be created for a skeleton element in review mode (https://github.com/opencv/cvat/pull/7362)

Changed

• When the ORG_INVITATION_CONFIRM setting is enabled, organization invitations for existing users are no longer accepted automatically. Instead, the invitee can now review the invitation and choose to accept or decline it. (https://github.com/opencv/cvat/pull/7138)

• [Compose, Helm] Updated Clickhouse to version 23.11.* (https://github.com/opencv/cvat/pull/7268)

• Job queues are now stored in a dedicated Redis instance (https://github.com/opencv/cvat/pull/7245)

Removed

• PermissionDenied error thrown before OPA call in case if user is not a member of organization (https://github.com/opencv/cvat/pull/7259)

Fixed

• Can not input Chinese correctly in text attributes on objects sidebar (https://github.com/opencv/cvat/pull/6916)

• Restored Compose file compatibility with Docker Compose 2.17.0 and earlier (https://github.com/opencv/cvat/pull/7170)

• Attaching GCS and AWS S3 buckets with dots in name (https://github.com/opencv/cvat/pull/7227)

• Annotation actions are applied to the objects from a ground truth job (https://github.com/opencv/cvat/pull/7281)

• Ground truth objects removed together with annotation objects when press "Remove annotations" in menu (https://github.com/opencv/cvat/pull/7281)

• Frame search by a filter is affected by ground truth annotations (https://github.com/opencv/cvat/pull/7281)

• Creating duplicating annotations when nginx throws 504 timeout status (workaround) (https://github.com/opencv/cvat/pull/7286)

• TIFF images are saved as JPEG images with .tif extension in original chunks (https://github.com/opencv/cvat/pull/7100)

• EXIF rotated TIFF images are handled incorrectly (https://github.com/opencv/cvat/pull/7100)

• RQ Scheduler launch, broken in PR 7245 (https://github.com/opencv/cvat/pull/7293)

• UI crashes if user highligts conflict related to annotations hidden by a filter (https://github.com/opencv/cvat/pull/7299)

• Annotations conflicts are not highligted properly on the first frame of a job (https://github.com/opencv/cvat/pull/7299)

• Error message Edge's nodeFrom ${dataNodeFrom} or nodeTo ${dataNodeTo} do not to refer to any node when upload a file with some abscent skeleton nodes (https://github.com/opencv/cvat/pull/7302)

• Wrong context menu position in skeleton configurator (Firefox only) (https://github.com/opencv/cvat/pull/7302)

• Fixed console error (Error: <rect> attribute width: A negative value is not valid appearing when skeleton with all outside elements is created (https://github.com/opencv/cvat/pull/7302)

• Updating cloud storage attached to CVAT using Azure connection string (https://github.com/opencv/cvat/pull/7336)

Added

• Introduced CVAT actions. Actions allow performing different predefined scenarios on annotations automatically (e.g. shape converters) (https://github.com/opencv/cvat/pull/7172)

• The UI will now retry requests that were rejected due to rate limiting (https://github.com/opencv/cvat/pull/7216)

Changed

• Update nvidia/cuda image version from 11.7.0 to 11.7.1 in transt serverless function. (https://github.com/opencv/cvat/pull/7124)

• [Helm] Allow pre-release versions in kubernetes requirement to include AWS EKS versions (https://github.com/opencv/cvat/pull/7183)

• GPU versions of serverless functions now use the latest-gpu Docker tag rather than latest (https://github.com/opencv/cvat/pull/7215)

• [Compose, Helm] Downgraded KeyDB to 6.3.2 (https://github.com/opencv/cvat/pull/7236)

Fixed

• The GPU version of the YOLOv7 serverless function not actually using the GPU (https://github.com/opencv/cvat/pull/6940)

• It is now possible to create Ground Truth jobs containing all frames in the task (https://github.com/opencv/cvat/pull/7126)

• Incorrect Ground Truth chunks saving (https://github.com/opencv/cvat/pull/7126)

• Reset source/target storage if related cloud storage has been deleted (https://github.com/opencv/cvat/pull/6801)

• Prevent possible cyclic dependencies when enqueuing a rq job when ONE_RUNNING_JOB_IN_QUEUE_PER_USER is used (https://github.com/opencv/cvat/pull/7139)

• Enqueue deferred jobs when their dependencies are moved to the failed job registry due to AbandonedJobError (https://github.com/opencv/cvat/pull/7139)

• Reduce the number of requests to the server for task details (https://github.com/opencv/cvat/pull/7167)

• Shape settings opacity and selected opacity reset on each frame change (https://github.com/opencv/cvat/pull/7186)

• Server error in list quality settings API, when called in an org (https://github.com/opencv/cvat/pull/7190)

• Incorrect handling of the hidden points in skeletons in quality comparisons (https://github.com/opencv/cvat/pull/7191)

• [Helm] Fixed installing Traefik Middleware even if Traefik is disabled in the values (https://github.com/opencv/cvat/pull/7184)

• Error code 500 when send change:frame event without duration. (https://github.com/opencv/cvat/pull/7211)

• Added workaround for corrupted cached chunks (https://github.com/opencv/cvat/pull/7243, https://github.com/opencv/cvat/pull/7251)

This release has changes only in the Enterprise version.

Added

• CVAT now supports serverless Nuclio functions that return skeleton annotations. We've added a keypoint detector that supports skeletons for the following classes: body, head, foot, and hands. Deployment command: ./deploy_cpu.sh pytorch/mmpose/hrnet32/nuclio/ (https://github.com/opencv/cvat/pull/7033)

• Implemented a feature that allows slicing one polygon/mask shape into two parts (https://github.com/opencv/cvat/pull/7084)

• Implemented a feature that allows joining several masks into a single one (https://github.com/opencv/cvat/pull/7084)

• [Helm] Introduced values that apply to all backend deployments/jobs (https://github.com/opencv/cvat/pull/7148)

Changed

• The "use cache" option on the server is now ignored when creating a task with cloud storage data (https://github.com/opencv/cvat/pull/7087)

• The Docker Compose file and Helm chart have been updated to enable Traefik access logs by default and change the log format to JSON (https://github.com/opencv/cvat/pull/7109)

• [Helm] The PersistentVolumeClaim for the volume used to hold application data is now retained after uninstall (https://github.com/opencv/cvat/pull/7123)

• [Helm] All backend-related deployments now use cvat-app as the value for the app label (https://github.com/opencv/cvat/pull/7127)

• [Helm] The minimum compatible Kubernetes version is now 1.19.0 (https://github.com/opencv/cvat/pull/7132)

• [Helm] The CVAT hostname can now be configured with ingress.hostname option (https://github.com/opencv/cvat/pull/7132)

• [Helm] The ingress.tls configuration has been reworked. (https://github.com/opencv/cvat/pull/7132)

• [Helm] The Traefik subchart updated to 25.0.0 (appVersion v2.10.5) (https://github.com/opencv/cvat/pull/7132)

• [Docker Compose] Traefik updated to v2.10.* (https://github.com/opencv/cvat/pull/7150)

Removed

• Support for V1 cloudstorages/id/content endpoint (https://github.com/opencv/cvat/pull/6946)

• [Helm] ingress.hosts has been removed, use ingress.hostname instead. (https://github.com/opencv/cvat/pull/7132)

Fixed

• Fixed a data race condition during GT job creation (https://github.com/opencv/cvat/pull/7096)

• Resolved an issue where the job state could not be changed multiple times without reloading the annotation view (https://github.com/opencv/cvat/pull/7158)

• Corrected an issue where compressed chunks did not utilize the Exif rotation tag (https://github.com/opencv/cvat/pull/7162)

• Minor styling issues on empty models page (https://github.com/opencv/cvat/pull/7164)

• Fixed minor issue when brush marker is appended to a final mask (https://github.com/opencv/cvat/pull/7168)

Fixed

• OpenCV runtime initialization (https://github.com/opencv/cvat/pull/7101)

Added

• Support for default bucket prefix (https://github.com/opencv/cvat/pull/6943)

• Search for cloud storage and share files (https://github.com/opencv/cvat/pull/6943)

• Ability to limit one user to one task at a time (https://github.com/opencv/cvat/pull/6975)

• Support for using an external database in a Docker Compose-based deployment (https://github.com/opencv/cvat/pull/7055)

Changed

• Migrated to rq 1.15.1 (https://github.com/opencv/cvat/pull/6975)

• Compressed sequental change:frame events into one (https://github.com/opencv/cvat/pull/7048)

• Create a local session for AWS S3 client instead of using the default global one (https://github.com/opencv/cvat/pull/7067)

• Improved performance of chunk preparation when creating tasks (https://github.com/opencv/cvat/pull/7081)

Fixed

• Race condition in a task data upload request, which may lead to problems with task creation in some specific cases, such as multiple identical data requests at the same time (https://github.com/opencv/cvat/pull/7025)

• Bug with viewing dependent RQ jobs for downloading resources from cloud storage when file path contains sub-directories. This is relevant for admins that can view detailed information about RQ queues. (https://github.com/opencv/cvat/pull/6975)

• OpenCV.js memory leak with TrackerMIL (https://github.com/opencv/cvat/pull/7032)

• Can't deploy detectron serverless function (https://github.com/opencv/cvat/pull/7047)

• A mask becomes visible even if hidden after changing opacity level (https://github.com/opencv/cvat/pull/7060)

• There is no switcher to personal workspace if an organization request failed (https://github.com/opencv/cvat/pull/7063)

Added

• A new feature allowing users to invite others to the organization via email. (https://github.com/opencv/cvat/pull/6901)

• [SDK] In the SDK, a parameter has been introduced to TaskDataset which enables the option to disable annotation loading (https://github.com/opencv/cvat/pull/7019)

• A test has been incorporated for retrieving bucket content in cases where the bucket includes manually created directories. (https://github.com/opencv/cvat/pull/7018)

Changed

• The maximum length of the secret access key has been increased to 64 characters. (https://github.com/opencv/cvat/pull/6701)

• The client will no longer load all organizations upon start (https://github.com/opencv/cvat/pull/7004)

• The default value for Zookeeper from the Clickhouse subchart has been set to disabled. (https://github.com/opencv/cvat/pull/7003)

Removed

• The endpoints /api/projects, /api/tasks, and /api/jobs will no longer return information regarding the count of labels. This information was complicating SQL queries, making them hard to optimize. Instead, use /api/labels?task_id=tid or /api/labels?project_id=pid. (https://github.com/opencv/cvat/pull/6918)

Fixed

• Issues causing potential double-sized file writes during task data uploading have been addressed. (https://github.com/opencv/cvat/pull/6952)

• Issues encountered when retrieving CS content from GCS buckets containing manually created directories have been resolved. (https://github.com/opencv/cvat/pull/7006)

• [SDK] In the SDK, cvat_sdk.auto_annotation.annotate_task has been optimized to avoid unnecessary fetching of existing annotations. (https://github.com/opencv/cvat/pull/7019)

• The project/task/job update time is now correctly modified upon label updates. (https://github.com/opencv/cvat/pull/6958)

Changed

• Enabled nginx proxy buffering
  (https://github.com/opencv/cvat/pull/6991)

• Helm: set memory request for keydb
  (https://github.com/opencv/cvat/pull/6945)

• Supervisord (https://github.com/opencv/cvat/pull/6945):

  • added autorestart=true option for all workers
  • unified program names to use dashes as delimiter instead of mixed '_' and '-'
  • minor improvements to supervisor configurations

Removed

• Removed gitter link from about modal
  (https://github.com/opencv/cvat/pull/7002)

Fixed

• Persist image filters across jobs
  (https://github.com/opencv/cvat/pull/6953)

• Splitting skeleton tracks on jobs
  (https://github.com/opencv/cvat/pull/6968)

• Uploading skeleton tracks in COCO Keypoints format
  (https://github.com/opencv/cvat/pull/6969)

• Fixed Siammask tracker error on grayscale images
  (https://github.com/opencv/cvat/pull/6982)

• Fixed memory leak on client side when event listener was not removed together with its context
  (https://github.com/opencv/cvat/pull/6984)

• Fixed crash related to issue tries to mount to not existing parent
  (https://github.com/opencv/cvat/pull/6977)

• Added 'notranslate' markers to avoid issues caused by extension translators
  (https://github.com/opencv/cvat/pull/6993)

• Getting CS content when S3 bucket contains manually created directories
  (https://github.com/opencv/cvat/pull/6997)

• Optimized huge memory consumption when working with masks in the interface
  (https://github.com/opencv/cvat/pull/6996)

Security

• Security upgrade opencv-python-headless from 4.5.5.62 to 4.8.1.78
  (https://github.com/opencv/cvat/pull/6931)

• Added X-Frame-Options: deny
  (https://github.com/opencv/cvat/pull/6992)

New Contributors

• @WilliamKyle made their first contribution in https://github.com/opencv/cvat/pull/6961

Full Changelog: https://github.com/opencv/cvat/compare/v2.7.5...v2.7.6

• Temporary workaround to fix corrupted zip file
  (https://github.com/opencv/cvat/pull/6965)

Full Changelog: https://github.com/opencv/cvat/compare/v2.7.4...v2.7.5

Added

• The latest comment displayed in issues sidebar (https://github.com/opencv/cvat/pull/6937)

Fixed

• It was not possible to copy issue comment from issue dialog (https://github.com/opencv/cvat/pull/6937)

Security

• Update Grafana from 9.3.6 to 10.1.2

Full Changelog: https://github.com/opencv/cvat/compare/v2.7.3...v2.7.4

Added

• New , form-based Issue templates for Github repository

Removed

• Functionality for synchronizing a task with a Git repository
  (https://github.com/opencv/cvat/pull/6904)

Fixed

• Incorrectly determined video frame count when the video contains an MP4 edit list
  (https://github.com/opencv/cvat/pull/6929)
• Internal server error when retrieving data from CS and cache=True
  (https://github.com/opencv/cvat/pull/6932)

Security

• Security upgrade Pillow from 9.3.0 to 10.0.1
  (https://github.com/opencv/cvat/pull/6907)
• Security update cryptography from 41.0.3 to 41.0.4
  (https://github.com/opencv/cvat/pull/6914)

New Contributors

• @AnnaPetrovicheva made their first contribution in https://github.com/opencv/cvat/pull/6909
• @cuong-ts made their first contribution in https://github.com/opencv/cvat/pull/6762

Full Changelog: https://github.com/opencv/cvat/compare/v2.7.2...v2.7.3