Reproduce “DEMONet: Underwater Acoustic Target Recognition based on Multi-Expert Network and Cross-Temporal Variational Autoencoder” 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

Yuan Xie, Xiaowei Zhang, Jiawei Ren, Ji Xu

DEMONet uses DEMON spectra to route underwater ship sounds to specialist experts, while spectrogram-processing experts classify ships. It reconstructs DEMON spectra from different periods with a cross-temporal VAE, then trains the router and recognition network. On DeepShip, its five-expert system reaches 80.45 ± 0.67% segment-level accuracy, compared with 77.15 ± 1.16% for the baseline; on ShipsEar, it falls below the strongest listed method.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.