从音频或视频文件中识别鸟叫声对应的鸟种,基于康奈尔实验室 BirdNET v2.4 声学模型。自动抽取音轨(无需系统 ffmpeg),支持地理/季节校正,可生成语谱图核验。首次使用需联网下载约 50MB 模型。Identify bird species from an audio OR video file using the BirdNET v2.4 acoustic model. Use when the user provides a media file (mp4/mov/wav/mp3/m4a/etc.) and asks "what bird is this", "识别鸟叫/鸟种", "分辨这是什么鸟", "音视频鸟鸣识别", or wants acoustic bird-call identification. Auto-extracts the audio track (no system ffmpeg required), runs TFLite inference directly (bypassing the buggy high-level birdnet API), applies optional geographic/seasonal filtering, and can generate a spectrogram for verification. Fully portable — no hardcoded machine paths.
从音频或视频文件中识别鸟叫声对应的鸟种,基于康奈尔实验室 BirdNET v2.4 声学模型。自动抽取音轨(无需系统 ffmpeg),支持地理/季节校正,可生成语谱图核验。首次使用需联网下载约 50MB 模型。Identify bird species from an audio OR video file using the BirdNET v2.4 acoustic model. Use when the user provides a media file (mp4/mov/wav/mp3/m4a/etc.) and asks "what bird is this", "识别鸟叫/鸟种", "分辨这是什么鸟", "音视频鸟鸣识别", or wants acoustic bird-call identification. Auto-extracts the audio track (no system ffmpeg required), runs TFLite inference directly (bypassing the buggy high-level birdnet API), applies optional geographic/seasonal filtering, and can generate a spectrogram for verification. Fully portable — no hardcoded machine paths.