BSDR: A Data-Efficient Deep Learning-Based Hyperspectral Band Selection Algorithm Using Discrete Relaxation
Mohammad Rahman, Shyh Wei Teng, Manzur Murshed, Manoranjan Paul, David Brennan · 2024
Read BSDR: A Data-Efficient Deep Learning-Based Hyperspectral Band Selection Algorithm Using Discrete Relaxation by Mohammad Rahman, Shyh Wei Teng, Manzur Murshed, Manoranjan Paul, David Brennan (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11644938.txt.
Description
Mohammad Rahman and co-authors present a 2024 study of a lean deep learning method that picks the most useful colour bands from hyperspectral images. Such images carry hundreds of bands, most of them dead weight. Attention based pickers work but are heavy and data hungry.
Their method learns only the target bands, far fewer numbers, and turns the pick or skip choice into a smooth dial that gradient search can turn. On three benchmark sets it beat attention and classic pickers on accuracy for both sorting and number tasks, while running in a small fraction of the time. Less data, less compute, better picks.
Catalog record
| CREATOR | Mohammad Rahman, Shyh Wei Teng, Manzur Murshed, Manoranjan Paul, David Brennan |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |