| --- |
| license: mit |
| task: image-to-parameter |
| tags: |
| - solar-system |
| - sexagesimal |
| - time-lapse |
| - webxos |
| - multi-modal |
| - astronomy |
| - 360-panorama |
| - equirectangular |
| datasets: |
| - webxOS/solar-time-360 |
| --- |
| |
| # webXOS Solar Time Studio – 360° Panorama Dataset Sampler |
|
|
| This dataset sample contains **equirectangular 360° panoramas** of a synthetic solar system, with planetary positions encoded in base‑60 (sexagesimal) time. Each sample includes a time-lapse sequence of full 360° panoramic images and associated orbital parameters. |
|
|
| The SOLAR TIME GENERATOR by webXOS is available to download in the /generator/ folder of this repo. |
|
|
| ## Sample Structure |
| - `panoramas/`: equirectangular PNG panoramas named `pano_sample_XXX_frame_YYY.png` |
| - `metadata.json`: JSON array with per-frame entries including: |
| - `sample_id`, `frame_id` |
| - `time_sexagesimal`: HH:MM:SS in base‑60 |
| - `time_total_seconds`: integer seconds |
| - `planet_states`: angles, distances, positions for all 8 planets |
| - `image_file`: relative path to PNG |
| - `pano_resolution`: width of equirectangular panorama |
|
|
| ## Sample Usage |
| Ideal for multi-modal model training (image + parameter regression), physics‑inspired ML, or time‑series forecasting with celestial mechanics. The 360° panoramas provide full scene context for data study. |
|
|
| ## Sample Parameters |
| - 8 planets + Sun |
| - Base‑60 time encoding |
| - Configurable time steps and frame rates |
| - Equirectangular projection (2:1 aspect ratio) |
|
|
| |
