Reproduce “A Hierarchical Method for Environmental Sounds Recognition in Complex Acoustic Scenes” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zhouwen Li, Haijun Lin, Bowen Lv, Yupeng Chen
FTANet separates time and frequency features with convolution branches and channel recalibration, then routes recordings through a broad sound category to a specialist classifier. On a 25-class test, two-stage FTANet raises accuracy from 71.32% to 81.49% and Macro-F1 from 68.51% to 79.58%. F1 improves for 20 of 25 classes, while reported inference time increases from 0.2170 s to 1.2390 s.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.