{"success":true,"data":{"author":"Norwegian University of Science and Technology (Ustad, Logacjov, Trollebø, Thingstad, Vereijken, Bach, Mork); harmonized by Josef Heidler","bucketAccessType":"public","bucketName":"josefheidler/har_older-adults_2023-harth","bucketPrefix":"","category":"Physiology","cloudProvider":"huggingface","createdAt":"2026-11-22T01:09:18.291Z","description":"HAR70+ (Human Activity Recognition 70+) from NTNU in Trondheim: raw triaxial accelerometry of 18 older adults aged 70 to 95, five of whom used walking aids, recruited from a rehabilitation clinic. Each wore two Axivity AX3 accelerometers (plus or minus 8 g, 50 Hz) on the right thigh, about 10 cm above the kneecap, and on the lower back near the third lumbar vertebra during a roughly 40-minute semi-structured free-living session, while a trained observer logged activity in real time. Seven activity types are labelled: walking, shuffling, stair ascent, stair descent, standing, sitting and lying. The release has about 2.26 million samples and 12.6 labelled hours, all samples labelled. It is the training set for activity type classifiers for older adults and frail populations, gait and mobility research, and transfer from the younger-adult HARTH dataset. This Hugging Face copy harmonizes the NTNU release into one Parquet file per participant under harmonized/, with back and thigh x, y, z acceleration in g and the activity label for every sample; load with Hugging Face datasets or pandas. The UCI Machine Learning Repository entry for HAR70+ (DOI 10.24432/C5CW3D) states the CC BY 4.0 license. Cite Ustad et al., Sensors 23(1):130, 2023.","displayTitle":"HAR70+: Thigh and Lower-Back Accelerometry of 18 Adults Aged 70 to 95 with Observer-Annotated Daily Activities (NTNU)","huggingFaceUrl":"https://huggingface.co/datasets/josefheidler/har_older-adults_2023-harth","id":"f8dc00d9-5f7c-5bcc-81fc-6e294f736dd4","license":"CC-BY-4.0","licenseEvidenceUrl":"https://archive.ics.uci.edu/dataset/780","objectCountEstimate":20,"probeStatus":"ok","referenceUrl":"https://doi.org/10.3390/s23010130","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"josefheidler/har_older-adults_2023-harth\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"josefheidler/har_older-adults_2023-harth\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=josefheidler, license:cc-by-4.0, size_categories:1M<n<10M, format:parquet, downloads=255"],"format":"mixed","generatedAt":"2026-11-22T01:10:15.937Z","layout":{"example":"harmonized/agile-anteater.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"504152311504152415284c1502150012000012440e0000006167696c652d616e7465617465721500151a151e2c15c0b8021510150615061c3600280e6167696c652d616e746561746572180e6167696c652d616e74656174657211110000000d3004000000c0b8020101c0b802001500151a151e2c15c0b8021510150615061c3600280e6167696c652d616e746561746572180e6167696c652d616e74656174657211110000000d3004000000c0b8020101c0b802001500151a151e2c15c0b8021510150615061c3600280e6167696c652d616e746561746572180e6167696c652d616e74656174657211110000000d3004000000c0b8020101c0b802001500151a151e2c15c0b8021510150615061c3600280e6167696c652d616e746561746572180e6167696c652d616e74656174657211110000000d3004000000c0b8020101c0b802001500151a151e2c15c0b8021510150615061c3600280e6167696c652d616e746561746572180e6167696c652d616e74656174657211110000000d3004000000c0b8020101c0b8020015001516151a2c15a83c1510150615061c3600280e6167696c652d616e746561746572180e6167696c652d616e74656174657211110000000b2803000000a83c0101a83c00150415c0b66515c0a7434c15e8d60c1500120000a0db3224184ed07648be0500389c09080440ee090808603cd1","encoding":"hex","key":"harmonized/agile-anteater.parquet","truncatedAt":512}],"sampleObjects":["harmonized/agile-anteater.parquet","harmonized/bright-blackbird.parquet","harmonized/calm-cormorant.parquet","harmonized/deft-dormouse.parquet","harmonized/eager-eel.parquet"],"sidecars":[{"bytes":"---\npretty_name: \"HAR70+ — Older Adult Activity Recognition (Thigh + Back Accelerometry)\"\nlicense: cc-by-4.0\nlanguage:\n  - en\ntags:\n  - accelerometry\n  - wearable\n  - human-activity-recognition\n  - time-series-classification\n  - thigh-worn\n  - physical-activity\n  - movement-behaviour\n  - time-series\n  - older-adults\n  - free-living\ntask_categories:\n  - tabular-classification\n  - other\nsize_categories:\n  - 1M<n<10M\nconfigs:\n  - config_name: default\n    data_files:\n      - split: train\n        path: \"harmonized/*.parquet\"\n---\n\n# HAR70+ — Older Adult Activity Recognition (Thigh + Back Accelerometry)\n\nTwo Axivity AX3 accelerometers (right thigh and lower back, 50 Hz) worn during\nsemi-structured free-living sessions by 18 older adults (age 70–95) in Norway.\nHarmonized from the [NTNU HAR70+ release](https://github.com/ntnu-ai-lab/harth-ml-experiments/tree/main/har70plus)\ninto per-subject parquet — one row per raw sample. 18 participants, 2.26 M samples,\n100% labelled, 12.6 h labeled.\n\n## Source\n\n- **Paper**: Ustad, A., Logacjov, A., Trollebø, S. Ø., Thingstad, P., Vereijken, B.,\n  Bach, K., & Mork, P. J. (2023). Validation of an Activity Type Recognition Model\n  Classifying Daily Physical Behavior in Older Adults. *Sensors*, 23(1), 130.\n  https://doi.org/10.3390/s23010130\n- **Raw data**: https://github.com/ntnu-ai-lab/harth-ml-experiments/tree/main/har70plus\n- **License**: CC-BY-4.0\n\n## Protocol\n\n- **Participants**: 18 older adults, age 70–95 (five used walking aids), recruited from\n  a rehabilitation clinic in Trondheim, Norway.\n- **Sensor**: Axivity AX3 triaxial accelerometer (±8 g, 50 Hz), placed on the right\n  thigh (≈10 cm above the kneecap) and lower back (≈3rd lumbar vertebra), USB connector\n  pointing down.\n- **Protocol**: ≈40-minute semi-structured free-living session per participant.\n  Participants performed their usual daily activities while a trained observer accompanied\n  them and annotated behaviour in real time. Seven activity types were annotated: walking,\n  shuffling, stair climbing (ascending and descending), standing, sitting, and lying.\n\n### Ground-truth provenance\n\nActivities were annotated by a trained observer accompanying each participant during the\nsession using a mobile logging application. Timestamps were recorded manually. No video\nannotation was used; inter-rater reliability is not reported in the paper. All 2,259,597\nsamples are labeled — there are no unlabeled transition windows.\n\n## Schema\n\nAll timestamps are tz-aware **Europe/Oslo**.\n\n| column       | dtype                       | notes                                                                      |\n|--------------|-----------------------------|----------------------------------------------------------------------------|\n| subject      | string                      | participant alias (e.g. `agile-anteater`)                                  |\n| timestamp    | datetime64[ns, Europe/Oslo] | real wall-clock time of the recording, tz-aware Europe/Oslo                |\n| thigh_acc_x  | float64                     | right-thigh acceleration, corrected frame (see Axis orientation), units = g |\n| thigh_acc_y  | float64                     | right-thigh acceleration, corrected frame, units = g                        |\n| thigh_acc_z  | float64                     | right-thigh acceleration, corrected frame, units = g                        |\n| back_acc_x   | float64                     | lower-back acceleration, corrected frame (see Axis orientation), units = g  |\n| back_acc_y   | float64                     | lower-back acceleration, corrected frame, units = g                         |\n| back_acc_z   | float64                     | lower-back acceleration, corrected frame, units = g                         |\n| label        | string (nullable)           | activity base (`walk`, `shuffle`, `stairs`, `stand`, `sit`, `lie`)          |\n| variant      | string (nullable)           | direction modifier (`ascending`, `descending`); null for most activities    |\n\n**Example rows**\n\n```\nsubject          timestamp                       t","encoding":"utf8","key":"README.md","name":"README.md","size":7989,"truncated":true}],"sourceObjectKey":"harmonized/agile-anteater.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"josefheidler/har_older-adults_2023-harth\")\nprint(ds)","starterSamplePath":"harmonized/agile-anteater.parquet","totalObjects":20},"summary":"HAR70+ (Human Activity Recognition 70+) from NTNU in Trondheim: raw triaxial accelerometry of 18 older adults aged 70 to 95, five of whom used walking aids, recruited from a…","title":"HAR70+: Thigh and Lower-Back Accelerometry of 18 Adults Aged 70 to 95 with Observer-Annotated Daily Activities (NTNU)","updatedAt":"2026-11-22T04:00:16.680Z","visibility":"public","websiteUrl":"https://github.com/ntnu-ai-lab/harth-ml-experiments"}}