# ranjeet-2/antariksh-jepa architecture

> Architecture diagram of the GitHub repository ranjeet-2/antariksh-jepa, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-20). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/ranjeet-2/antariksh-jepa
- Repository: https://github.com/ranjeet-2/antariksh-jepa

## Overview

Antariksh-JEPA is a cross-modal satellite-image retrieval app: a user supplies an optical or SAR query, the app prepares and loads held-out imagery and matching model/index artifacts, then presents ranked matches and spatial context. The sampled runtime is concentrated in `Web_app/app.py`; `Web_app/asset_cache.py` implements test-tile extraction. The README documents broader training and evaluation methodology, but those workflows live mainly in notebooks whose implementation was not sampled, so they are described only as documented capabilities rather than linked into the runtime flow.

Main entry points: [`Web_app/app.py`](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py) (Streamlit dashboard), [`Web_app/asset_cache.py`](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/asset_cache.py) (Test tile cache)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_query["Query experience"]
  node_dashboard["Streamlit dashboard<br/>[app.py]"]
  node_query_input["Image query<br/>[app.py]"]
end

subgraph group_data["Data and assets"]
  node_test_data[("Test imagery<br/>[app.py]")]
  node_remote_dataset[("Dataset repository")]
  node_tile_cache[("Test tile cache<br/>[asset_cache.py]")]
  node_run_assets[("Model run assets<br/>[app.py]")]
end

subgraph group_retrieval["Retrieval engine"]
  node_checkpoint["Contrastive checkpoint<br/>[app.py]"]
  node_retrieval_model["Antariksh-JEPA model<br/>[app.py]"]
  node_binary_indexes[("Binary FAISS indexes<br/>[app.py]")]
  node_gallery_metadata[("Gallery metadata<br/>[app.py]")]
end

subgraph group_results["Results and mapping"]
  node_ranked_matches["Ranked matches<br/>[app.py]"]
  node_previews["Image previews<br/>[app.py]"]
  node_map_tiles["Esri map tiles"]
  node_spatial_view["Spatial results view<br/>[app.py]"]
end

node_user(("Analyst"))

node_user -->|"submits query"| node_dashboard
node_dashboard -->|"accepts image"| node_query_input
node_dashboard -->|"downloads data"| node_remote_dataset
node_dashboard -->|"extracts tiles"| node_tile_cache
node_tile_cache -->|"writes tiles"| node_test_data
node_dashboard -->|"loads imagery"| node_test_data
node_dashboard -->|"loads bundle"| node_run_assets
node_run_assets -->|"includes checkpoint"| node_checkpoint
node_run_assets -->|"includes indexes"| node_binary_indexes
node_run_assets -->|"includes metadata"| node_gallery_metadata
node_dashboard -->|"instantiates model"| node_retrieval_model
node_checkpoint -->|"loads weights"| node_retrieval_model
node_dashboard -->|"loads indexes"| node_binary_indexes
node_dashboard -->|"loads metadata"| node_gallery_metadata
node_dashboard -->|"shows top matches"| node_ranked_matches
node_dashboard -->|"renders imagery"| node_previews
node_dashboard -->|"shows locations"| node_spatial_view
node_spatial_view -->|"requests tiles"| node_map_tiles

click node_dashboard "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_query_input "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_test_data "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_tile_cache "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/asset_cache.py"
click node_run_assets "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_checkpoint "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_retrieval_model "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_binary_indexes "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_gallery_metadata "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_ranked_matches "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_previews "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"
click node_spatial_view "https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_dashboard,node_query_input toneBlue
class node_test_data,node_remote_dataset,node_tile_cache,node_run_assets toneAmber
class node_checkpoint,node_retrieval_model,node_binary_indexes,node_gallery_metadata toneMint
class node_ranked_matches,node_previews,node_map_tiles,node_spatial_view toneRose
class node_user toneIndigo
```

## Components

### Query experience

- [Streamlit dashboard](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`
- [Image query](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`

### Data and assets

- [Test imagery](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`
- **Dataset repository**
- [Test tile cache](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/asset_cache.py): `Web_app/asset_cache.py`
- [Model run assets](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`

### Retrieval engine

- [Contrastive checkpoint](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`
- [Antariksh-JEPA model](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`
- [Binary FAISS indexes](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`
- [Gallery metadata](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`

### Results and mapping

- [Ranked matches](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`
- [Image previews](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`
- **Esri map tiles**
- [Spatial results view](https://github.com/ranjeet-2/antariksh-jepa/blob/main/Web_app/app.py): `Web_app/app.py`

### Other components

- **Analyst**

## Connections

- Analyst → Streamlit dashboard: submits query (evidence: `README.md`)
- Streamlit dashboard → Image query: accepts image (evidence: `Web_app/app.py`)
- Streamlit dashboard → Dataset repository: downloads data (evidence: `README.md`)
- Streamlit dashboard → Test tile cache: extracts tiles (evidence: `Web_app/app.py`)
- Test tile cache → Test imagery: writes tiles (evidence: `Web_app/asset_cache.py`)
- Streamlit dashboard → Test imagery: loads imagery (evidence: `Web_app/app.py`)
- Streamlit dashboard → Model run assets: loads bundle (evidence: `Web_app/app.py`)
- Model run assets → Contrastive checkpoint: includes checkpoint (evidence: `Web_app/app.py`)
- Model run assets → Binary FAISS indexes: includes indexes (evidence: `Web_app/app.py`)
- Model run assets → Gallery metadata: includes metadata (evidence: `Web_app/app.py`)
- Streamlit dashboard → Antariksh-JEPA model: instantiates model (evidence: `Web_app/app.py`)
- Contrastive checkpoint → Antariksh-JEPA model: loads weights (evidence: `Web_app/app.py`)
- Streamlit dashboard → Binary FAISS indexes: loads indexes (evidence: `Web_app/app.py`)
- Streamlit dashboard → Gallery metadata: loads metadata (evidence: `Web_app/app.py`)
- Streamlit dashboard → Ranked matches: shows top matches (evidence: `README.md`)
- Streamlit dashboard → Image previews: renders imagery (evidence: `Web_app/app.py`)
- Streamlit dashboard → Spatial results view: shows locations (evidence: `Web_app/app.py`)
- Spatial results view → Esri map tiles: requests tiles (evidence: `Web_app/app.py`)
