Code and Data for: A fine-tuning workflow for CNN-based bat acoustic classification

This repository contains all data and code used in the analyses presented in the manuscript "A fine-tuning workflow for CNN-based bat acoustic classification". The materials provided here ensure transparency and reproducibility of the complete analytical workflow, from raw field recordings and training databases, through CNN training and classification, to the statistical analyses and final tables and figures reported in the article.本代码仓库包含了学术稿件《基于卷积神经网络（Convolutional Neural Network，CNN）的蝙蝠声学分类微调工作流》中各项分析所使用的全部数据与代码。本仓库提供的所有材料可保障完整分析流程的可透明性与可复现性，覆盖从原始野外录音与训练数据集出发，经CNN训练与分类环节，直至论文中呈现的统计分析、最终表格与插图的全流程。