Skip to content

Releases: deepfieldlabs/astroLens

AstroLens v1.2.0

24 Mar 04:48
@github-actions github-actions
v1.2.0
6dcaf84

Choose a tag to compare

View all tags
AstroLens v1.2.0 Latest
Latest

AstroLens v1.2.0

AI-Powered Astronomical Anomaly Detection -- autonomous sky survey analysis with self-correcting intelligence.

What's New in v1.1.0

  • Streaming Discovery: Fully autonomous multi-day pipeline that downloads, analyzes, and reports anomaly candidates 24/7
  • Self-Correcting Intelligence: Auto-adjusts thresholds, rebalances sources, recalibrates detection in real time
  • YOLO v1.1 Model: Fine-tuned transient detector achieving 99.5% mAP50 (up from 51.5%)
  • Daily HTML Reports: Automated charts, rankings, and trend analysis
  • Web Streaming Dashboard: Live monitoring with Chart.js visualizations
  • Desktop Streaming Panel: Start, stop, and monitor from the native UI

Validated in autonomous operation: 22,195 images analyzed across 5,471 sky regions, 3,541 anomaly candidates, 269 known objects independently recovered and confirmed against SIMBAD/NED (including SN 2014J, NGC 3690, SDSS J0252+0039).

Downloads

Platform File
macOS (Apple Silicon) AstroLens-v1.2.0-macos-arm64.zip
Windows (x64) AstroLens-v1.2.0-windows-x64.zip
Linux / All platforms Docker (see below)

Docker

docker pull ghcr.io/deepfieldlabs/astrolens:v1.2.0
docker run -p 8000:8000 -p 8080:8080 ghcr.io/deepfieldlabs/astrolens:v1.2.0

Requirements

  • Python 3.10+ (or Docker)
  • 8GB+ RAM recommended
  • GPU optional (CUDA / Apple MPS auto-detected)

Full documentation: Website | README | Wiki

Assets 4
Loading

AstroLens v1.1.0

17 Feb 02:45
@github-actions github-actions
v1.1.0
e8d9a74

Choose a tag to compare

View all tags
AstroLens v1.1.0

AstroLens v1.1.0

AI-Powered Astronomical Anomaly Detection -- autonomous sky survey analysis with self-correcting intelligence.

What's New in v1.1.0

  • Streaming Discovery: Fully autonomous multi-day pipeline that downloads, analyzes, and reports anomaly candidates 24/7
  • Self-Correcting Intelligence: Auto-adjusts thresholds, rebalances sources, recalibrates detection in real time
  • YOLO v1.1 Model: Fine-tuned transient detector achieving 99.5% mAP50 (up from 51.5%)
  • Daily HTML Reports: Automated charts, rankings, and trend analysis
  • Web Streaming Dashboard: Live monitoring with Chart.js visualizations
  • Desktop Streaming Panel: Start, stop, and monitor from the native UI

Validated in a 3-day autonomous run: 20,997 images analyzed, 3,458 anomaly candidates, known objects independently recovered (SN 2014J, NGC 3690, SDSS J0252+0039).

Downloads

Platform File
macOS (Apple Silicon) AstroLens-v1.1.0-macos-arm64.zip
Windows (x64) AstroLens-v1.1.0-windows-x64.zip
Linux / All platforms Docker (see below)

Docker

docker pull ghcr.io/samantaba/astrolens:v1.1.0
docker run -p 8000:8000 -p 8080:8080 ghcr.io/samantaba/astrolens:v1.1.0

Requirements

  • Python 3.10+ (or Docker)
  • 8GB+ RAM recommended
  • GPU optional (CUDA / Apple MPS auto-detected)

Full documentation: README | Wiki

Loading

AstroLens v1.0.0

08 Feb 12:08
@github-actions github-actions
v1.0.0
0fda0e4

Choose a tag to compare

View all tags
AstroLens v1.0.0

AstroLens v1.0.0

AI-Powered Galaxy Anomaly Discovery System.

Downloads

Platform File
macOS (Apple Silicon) AstroLens-v1.0.0-macos-arm64.zip
Windows (x64) AstroLens-v1.0.0-windows-x64.zip
Linux / All platforms Docker (see below)

Linux & Docker (Recommended)

docker pull ghcr.io/samantaba/astrolens:v1.0.0
docker run -p 8000:8000 -p 8080:8080 ghcr.io/samantaba/astrolens:v1.0.0
# Open http://localhost:8080

What's Included

  • Pre-trained YOLO transient detection model (no training required)
  • Galaxy morphology analysis (CAS, Gini-M20)
  • Web interface for browser-based access
  • Multi-source data pipeline (DECaLS, SDSS, Pan-STARRS)

Requirements

  • Docker (for container), OR
  • Python 3.10+ (for running from source)
  • 8GB+ RAM recommended
  • GPU optional (CUDA / Apple MPS auto-detected)

See README for full documentation.

Loading