MiteshPuthran/Speech-Emotion-Analyzer ? reverse-engineered prompt

Build me a speech emotion analyzer that can listen to a short audio clip and tell me the emotion, plus whether the voice sounds male or female. I want it to work on recorded audio files and also let me test it with live voice input from the microphone. Please train it with the kinds of speech emotion datasets used in the project, and show the usual audio visualizations too, like waveform and spectrogram, so I can see what the model is reading.

The app should extract features from the audio, run a classification model, and show the predicted label clearly, along with the real versus predicted results for test samples. If you need to, look up the current docs for any audio or deep learning libraries you use. Also include a simple way to save the trained model and reuse it later, and make sure it can handle the common emotion labels like angry, calm, fearful, happy, and sad.