Human Activity Recognition Data Migration

Last updated on 


Introducing the Human Activity Recognition Data Migration

1. Check raw data

It seems that a new time column needs to be created, and preprocessing of the label 'Activity' is required.

2. Set the Preprocessing Plan

  • Data Cleaning.
    • Generate timestamps with a 1-minute interval.
    • Convert label values.
  • Convert to the format: Name, Time, Value & Convert to UTC time.

Data Cleaning

  • Generate timestamps with a 1-minute interval.
  • Convert label values.

Convert to the format: Name, Time, Value & Convert to UTC time

Once the data frame is restructured as shown in the image below, it will be ready for upload to Machbase Neo.

3. Data Upload

machbase-neo shell import --input ./datahub-2025-4-human-activity.csv.gz --compress gzip --header --method append --timeformat ns activity

Check the entire code.

datahub/dataset/2025/04.Human Activity Recognition/conv/convert.py at main · machbase/datahub
All Industrial IoT DataHub with data visualization and AI source - machbase/datahub
GitHubmachbase

4. Check the results after uploading

Output when executing the following code in the Machbase Neo internal shell.

select * from v$activity_stat;

※ Subsequent link to the AI training process: Human Activity Recognition Data

Back to Top