{"success":true,"data":{"author":"Imageomics Institute (Jenna Kline)","bucketAccessType":"public","bucketName":"imageomics/drone-maneuver-clips","bucketPrefix":"","category":"Life Sciences","cloudProvider":"huggingface","createdAt":"2026-11-19T15:09:27.743Z","description":"A benchmark of 41 six-second drone video clips (180 frames at 30 fps) cut from 20 KABR (Kenyan Animal Behavior Recognition) videos recorded over 17 survey sessions at the Mpala Research Centre, Kenya, showing giraffes, plains zebras and Grevy's zebras. Each clip is indexed by the autonomous-flight maneuver it can test (launch, follow, behavior-adaptive, SoI-aware) and ships a per-frame, per-track label table with species, expert behavior label, vigilance flag, 8-class viewpoint pose where ground truth exists, bounding boxes with size class, occlusion flags and drone latitude, longitude and altitude, plus per-frame ground-truth drone actions for each maneuver. Layout: catalog/ (video_index.csv, clip_index.csv, coverage report, pose audit) and clips/<clip_id>/ with clip.mp4, labels.csv and maneuver_labels.csv; source videos resolve to the KABR FAIR2 release by event ID. Supports detection, tracking, behavior recognition and maneuver-conditioned evaluation of drone navigation policies. About 4.7 GB. The compilation is CC BY 4.0 by the Imageomics Institute (curated by Jenna Kline); also cite KABR and KABR-poses.","displayTitle":"KABR Drone Maneuver Test-Clip Library: 41 Wildlife Drone Clips Labeled per Frame for Autonomous Flight Policy Evaluation","huggingFaceUrl":"https://huggingface.co/datasets/imageomics/drone-maneuver-clips","id":"1ed66eec-77ea-51d0-9e39-ec40c3a6894f","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/imageomics/drone-maneuver-clips","objectCountEstimate":130,"probeStatus":"ok","referenceUrl":"https://huggingface.co/datasets/imageomics/drone-maneuver-clips","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"imageomics/drone-maneuver-clips\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"imageomics/drone-maneuver-clips\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=imageomics, license:cc-by-4.0, size_categories:n<1K, downloads=270"],"format":"mixed","generatedAt":"2026-11-19T15:21:53.102Z","layout":{"example":"catalog/clip_index.csv","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"clip_id,video_id,session_id,date_prefix,start_frame,end_frame,start_time,habitat,habitat_notes,habitat_provenance,herd_size,pose_frames_covered,fair2_video_eventID,fair2_session_eventID,present_frac,n_tracks_max,multi_frac,max_area_frac,vigilant_to_calm_transition,species_set,behaviours_present,bbox_size_classes,suitable_maneuvers,pose_set\n11_01_23-DJI_0979_007485,DJI_0979,KABR-2023:11_01_23_session_2,11_01_23,7485,7665,249.5,mixed,Open grassy habitat with some scattered bushes,session_event,3,0,KABR-2023:1","encoding":"utf8","key":"catalog/clip_index.csv","truncatedAt":512}],"sampleObjects":["catalog/clip_index.csv","catalog/coverage_report.md","catalog/pose_audit.csv","catalog/video_index.csv","clips/11_01_23-DJI_0979_007485/clip.mp4"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\nlanguage:\n- en\npretty_name: \"Drone Maneuver Test-Clip Library\"\ntask_categories:\n- object-detection\n- video-classification\ntags:\n- biology\n- image\n- animals\n- CV\n- drone\n- UAV\n- KABR\n- zebra\n- giraffe\n- behavior\n- pose\n- animal-tracking\n- Mpala Research Centre\nsize_categories:\n- n<1K\ndescription: \"A library of 6-second aerial drone clips derived from the KABR dataset (Mpala Research Centre, Kenya), each indexed by the autonomous-flight maneuver it is suitable for testing, with per-frame-per-track labels (bounding box, species, behaviour, persistent track id, ground-truth pose where available, telemetry).\"\n---\n\n# Dataset Card for Drone Maneuver Test-Clip Library\n\nA benchmark of **41 six-second drone clips** (180 frames @ 30 fps)\ncut from 20 KABR videos across 17 survey sessions at the Mpala Research Centre,\nKenya. Each clip is indexed by which autonomous-flight **maneuver** it can test\n(launch / follow / behavior-adaptive / SoI-aware) and ships with a per-frame-per-track label table.\nThe library is intended for evaluating drone navigation policies against real wildlife footage.\n\n## Dataset Details\n\n### Dataset Description\n\n- **Curated by:** Jenna Kline\n- **Language(s) (NLP):** en\n- **Homepage:** https://imageomics.github.io/wildlife-drone-maneuver/\n- **Paper:** in prep\n- **Related dataset:** [imageomics/KABR-mini-scene-raw-videos](https://huggingface.co/datasets/imageomics/KABR-mini-scene-raw-videos),\n  [imageomics/KABR-raw-videos](https://huggingface.co/datasets/imageomics/KABR-raw-videos)\n\nThis dataset repackages the KABR aerial behavior dataset into short, mix-and-match clips that each\nexercise a specific drone maneuver, so navigation policies can be benchmarked per-maneuver rather\nthan only end-to-end. Labels are aligned per frame per tracked individual.\n\n### Supported Tasks and Leaderboards\n\nObject detection / tracking, behaviour recognition, pose (viewpoint) estimation, and\n**maneuver-conditioned navigation-policy evaluation** (the primary intended use).\n\n## Dataset Structure\n\n```\nkabr_clips/\n    catalog/\n        video_index.csv      # per source video: frame summary + joined FAIR² metadata\n        clip_index.csv       # one row per clip (the master index)\n        coverage_report.md   # species x habitat x bbox-size x maneuver coverage\n        pose_audit.csv       # per-video GT-pose assignment audit\n    clips/\n        <clip_id>/\n            clip.mp4             # 6 s, 180 frames @ 30 fps\n            labels.csv           # one row per frame per track\n            maneuver_labels.csv  # per-frame ground-truth drone action, per maneuver\n    README.md\n```\n\nThe **source video** for each clip is referenced by its `fair2_video_eventID`, which resolves\nin the [imageomics/kabr-full-release](https://huggingface.co/datasets/imageomics/kabr-full-release)\nFAIR² dataset (`data/video_events.csv`); the session it belongs to is `fair2_session_eventID`\n(`data/session_events.csv`). The catalog carries no absolute filesystem paths.\n\n`clip_id = <date>-<video>_<start_frame>` (e.g. `18_01_2023_session_7-DJI_0070_000360`).\n\n### Data Fields\n\n**clips/<clip_id>/labels.csv** (one row per frame per track):\n- `clip_id`, `video_id`, `session_id`: identifiers; `session_id` is the KABR/FAIR² session event.\n- `frame_global`: absolute frame in the source video; `frame_local`: 0-based within the clip; `time_s`.\n- `track_id`: KABR mini-scene id, **persistent within the source video**.\n- `species`, `behavior`: KABR expert ground-truth labels; `vigilant`: behavior in {Head Up, Running, Trotting}.\n- `pose`, `pose_provenance`, `pose_match_score`: 8-class viewpoint where ground-truth exists (see Annotations); else empty.\n- `individual_id`: empty — global re-identification is future work (no ground truth exists).\n- `xtl,ytl,xbr,ybr`, `x_c,y_c,w,h`, `bbox_area_frac`, `bbox_size_class` (far/medium/close, relative to this survey).\n- `occluded`, `outside`: KABR annotation flags; `latitude`, `longitude`, `altitude`: drone telemetry; `date_time`.\n\n**catalog/clip_index.csv** (one row per clip): identifiers, frame range, `species_set`, `ha","encoding":"utf8","key":"README.md","name":"README.md","size":13942,"truncated":true}],"sourceObjectKey":"catalog/clip_index.csv","starterCell":"import pandas as pd\ndf = pd.read_csv(\"hf://datasets/imageomics/drone-maneuver-clips/catalog/clip_index.csv\")\ndf.head()","starterSamplePath":"catalog/clip_index.csv","totalObjects":130},"summary":"A benchmark of 41 six-second drone video clips (180 frames at 30 fps) cut from 20 KABR (Kenyan Animal Behavior Recognition) videos recorded over 17 survey sessions at the Mpala…","title":"KABR Drone Maneuver Test-Clip Library: 41 Wildlife Drone Clips Labeled per Frame for Autonomous Flight Policy Evaluation","updatedAt":"2026-11-19T19:17:36.343Z","visibility":"public","websiteUrl":"https://imageomics.github.io/wildlife-drone-maneuver/"}}