FiftyOne Timestamps

This plugin provides operators to compute datetime related fields based on samples' filepath or created_at date.

This enables you to filter your dataset by:

• datetime time range
• weekday where 0 is monday and 6 is sunday
• time where 6.5 would be 6am and 19.0 would be 7pm
• timeofday based on provided geolocation can be one of dawn, sunrise, morning, day, evening, sunset, dusk, night

Installation

fiftyone plugins download https://github.com/mmoollllee/fiftyone-timestamps/

Python SDK

You can use the compute operators from the Python SDK!

import fiftyone as fo
import fiftyone.operators as foo

dataset = fo.load_dataset("existing-dataset")

compute_timestamps = foo.get_operator("@mmoollllee/timestamps/compute_timestamps")

## Compute from filepath with custom regex pattern where filenames look like `image-2024-12-30_23-59-59.jpg`
compute_timestamps(dataset, source="filepath", regex=r".*([0-9]{4})-?([0-9]{2})-?([0-9]{2})_([0-9]{2})?-?([0-9]{2})?-?([0-9]{2})?.*?", timezone="Europe/Berlin")

## Compute from created_at
compute_timestamps(dataset, source="created_at", timezone="Europe/Berlin")

## If geo[lat, long] is set, `timeofday` will be computed with "sunrise", "dawn, "morning", "day", "evening", "sunset", "dusk", "night"
compute_timestamps(dataset, geo=[48.12345,9.12345], timezone="Europe/Berlin")

ToDos

• Implement timeofday field with Daytime / Nighttime / Morning / Evening by using suntime