The Palomar Transient Factory

on YouTube ↗54m

" summary

The Palomar Transient Factory builds a real-time pipeline that moves 250 MB/min (≈100 GB/night) from a 92 MP camera through image subtraction to a database, producing ~1.5 million candidates per night of which ~50,000 are asteroids and only 3–4 are new young supernovae; a random-forest classifier assigns a real bogus score to candidates and reduced the human load, Galaxy Zoo Supernova crowdsourcing filled gaps until automation improved, end-to-end processing drops below ten minutes and enables spectra within hours and 2,700+ spectroscopically confirmed supernovae.

AI-written summary · about

" chapters (22)

1. 0:00Introduction
2. 0:28Supernova
3. 5:40Victor Blanco
4. 10:41Palomar Transient Factory
5. 12:42Telescope
6. 13:38Resources
7. 14:09Pipeline
8. 14:54Networking
9. 15:55Database
10. 17:18Hand Scanning
11. 20:04Galaxy Zoo Supernova
12. 23:18Pipeline Automation
13. 27:57Machine Learning
14. 28:45Data
15. 30:25Flash Spectroscopy
16. 33:03Hydrogen Gas
17. 34:31gravitational waves
18. 38:11future problems
19. 39:23new subtractions
20. 39:58final slide
21. 50:47prospecting for strange
22. 52:19publishing