THULab/sensor_dataset

Forearm EMG Sensor Dataset (TsFile)

Apache TsFile version of `AnasElkhabbaz/sensor_dataset`.

Overview

Surface electromyography (EMG) recordings of the forearm, packaged in the LeRobot v2.1 layout (robot_type = None). Each episode is one resting-state recording trial of a subject, sampling three forearm muscle sensors over time.

• —Episodes: 32 recording trials.
• —Frames: 152,837 timesteps total, sampled at ~3.33 Hz (each trial ≈ 1500 s).
• —Tasks: 32, one per episode — each labelled <Subject>_<Sex>_<N>_Rest (e.g. Billy_Male_1_Rest). task_index equals episode_index.
• —Sensors: 3 forearm EMG channels — brachioradialis, flexorcarpiradialis, flexorcarpiulnaris.
• —Source format: LeRobot v2.1 (no action and no video columns).

Schema (TsFile structure)

• —Time (INT64, milliseconds) — per-frame timestamp, restarts at 0 for each episode (round(timestamp * 1000); ~300 ms apart at 3.33 Hz).
• —episode_index (TAG) — trial id (0–31). Query one trial with WHERE episode_index=0.
• —task_index (TAG) — task id (equals episode_index); maps to the subject / trial label in meta/tasks.jsonl.
• —frame_index (FIELD, INT64) — frame number within the episode.
• —sample_index (FIELD, INT64) — the source global index column.
• —observation_state_0 (FIELD, FLOAT) — brachioradialis EMG.
• —observation_state_1 (FIELD, FLOAT) — flexorcarpiradialis EMG.
• —observation_state_2 (FIELD, FLOAT) — flexorcarpiulnaris EMG.

The source timestamp column is dropped because it equals Time / 1000.

Usage

Read the .tsfile file with the Apache TsFile Java or Python SDK.

# pip install tsfile
from tsfile import TsFileReader
reader = TsFileReader("data/anaselkhabbaz_sensor_dataset.tsfile")
# query one trial: WHERE episode_index=0

Source & license

• —Original dataset: https://huggingface.co/datasets/AnasElkhabbaz/sensor_dataset
• —Author / publisher: AnasElkhabbaz
• —License: Apache-2.0 (inherited from the original dataset)