README.md

---
title: Maya4
emoji: 📊
colorFrom: indigo
colorTo: indigo
sdk: static
pinned: true
license: apache-2.0
thumbnail: >-
  https://cdn-uploads.huggingface.co/production/uploads/68a43a86eea45496edb28ba6/hJyITwdmrofMtEtX8qFmW.png
---

# 🏢 Maya4  

![Maya4 Logo](Maya4.png)


# 🌌 Maya4  

**Maya4** is a project dedicated to curating and providing **multi-level intermediate SAR representations** from **Sentinel-1** acquisitions, spanning the entire chain from **Level 0 to Level 1**.  

The name *Maya4* draws inspiration from the **Māyā veil** in philosophy, where reality is hidden behind successive layers—just as radar echoes undergo transformations before forming a final SAR image.  

---

## 🚀 Mission  

- 📡 Curate and structure **Sentinel-1 Stripmap SAR data**  
- 🧩 Provide **intermediate signal representations** for research & ML pre-training  
- ☁️ Enable scalable, **cloud-native access** in **Zarr format**  

---

## 📦 Pre-Training Dataset Spaces  

Maya4 provides **Pre-Training (PT) dataset** exposing intermediate SAR signal states in a modern  (Zarr Format):  

| 📂 Dataset Split | 📊 Contents | 🛰️ Acquisition Mode | 💾 Size |
|------------------|-------------|----------------------|---------|
| **pt1** | [LINK](https://huggingface.co/datasets/Maya4/pt1) | Stripmap | 17 TB |
| **pt2** | [LINK](https://huggingface.co/datasets/Maya4/pt2) | Stripmap | 17 TB |
| **pt3** | [LINK](https://huggingface.co/datasets/Maya4/pt3) | Stripmap | 17 TB |
| **pt4** | [LINK](https://huggingface.co/datasets/Maya4/pt4) | Stripmap | 17 TB |
| **Total** | — | — | **68 TB** |


### 📊 Processing Levels

A major characteristic of this dataset lies in its sharding multi-level organisation:

<img src="https://i.ibb.co/Wv7SXd4N/intermediates.jpg" alt="Intermediates" width="100%">


| 🌐 Level | 🔤 Abbrev. | 📝 Description | 🎯 Purpose / Value |
|----------|-----------|----------------|--------------------|
| 📡 **Raw** | raw | Unprocessed radar echoes as recorded by Sentinel-1 | Baseline data; allows full custom SAR processing |
| 🎚️ **Range Compressed** | rc | Echoes compressed in the range dimension via matched filtering | Improves SNR; isolates scatterers along range |
| 🎯 **Range Cell Migration Corrected** | rcmc | Motion-compensated echoes with corrected range migration | Preserves geometric fidelity; enables proper azimuth focusing |
| 🖼️ **Azimuth Compressed** | ac | Fully focused SAR image in slant-range geometry | Standard **Level-1 product**; interpretable imagery |



---


### 🌍 PT Use Cases  

The following use cases can be addressed with Maya4:

![🌌 Maya4 Dataset Cards](Maya4_cards.jpg)


---

## 🙏 Acknowledgements  

- 🛰️ Data provided by the **Copernicus Sentinel-1 mission (ESA)**  
- 🏢 Curated and maintained by the **Maya4 organization**