EcoNet-Multi: optimising deep learning frameworks for robust waste detection in diverse global environments
Alok Kumar Shukla, Shubhra Dwivedi · 2026
Read EcoNet-Multi: optimising deep learning frameworks for robust waste detection in diverse global environments by Alok Kumar Shukla, Shubhra Dwivedi (2026) 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/PMC13457377.txt.
Description
Alok Kumar Shukla and Shubhra Dwivedi propose a smarter way to tune the deep learning models sorting waste from photographs. Sorting models often ace common categories and stumble on rare ones, and tuning them by hand trades accuracy against balanced detection. Their 2026 paper, part of the EcoNet-Multi project, introduces an optimizer chasing both goals: fewer classification errors and a higher macro-F1 score.
The algorithm, SMOBBCS, blends the bat algorithm's velocity updates with cuckoo search's random long jumps, shifting from broad exploration to fine local search. A cheap k-nearest-neighbor stand-in screens candidate settings so the expensive real evaluations stay few. Tested against five rival optimizers on four waste datasets and four classifier types, the method averaged 92.71 percent accuracy and a macro-F1 of 0.9209, with strong precision and recall. The authors aim it at automated recycling and monitoring systems that must work across varied settings.
Catalog record
| CREATOR | Alok Kumar Shukla, Shubhra Dwivedi |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 64 KB |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |