Inference code?

[arianaa30]

Hi, Thanks for the nice work. I was trying to just use your model for inference. I looked at the notebook and copied the necessary parts for inference, but get error This LabelEncoder instance is not fitted yet. Can you help what is missing in this code?

import os

from keras import regularizers
import keras
from keras.callbacks import ModelCheckpoint
from keras.layers import Conv1D, MaxPooling1D, AveragePooling1D, Dense, Embedding, Input, Flatten, Dropout, Activation, LSTM
from keras.models import Model, Sequential, model_from_json
from keras.preprocessing import sequence
from keras.preprocessing.sequence import pad_sequences
from keras.preprocessing.text import Tokenizer
from keras.utils import to_categorical
import librosa
import librosa.display
from matplotlib.pyplot import specgram
from sklearn.metrics import confusion_matrix
from sklearn.preprocessing import LabelEncoder

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import tensorflow as tf

opt = keras.optimizers.rmsprop(lr=0.00001, decay=1e-6)
lb = LabelEncoder()

json_file = open('model.json', 'r')
loaded_model_json = json_file.read()
json_file.close()
loaded_model = model_from_json(loaded_model_json)
# load weights into new model
loaded_model.load_weights("saved_models/Emotion_Voice_Detection_Model.h5")
print("Loaded model from disk")

X, sample_rate = librosa.load('h04.wav', res_type='kaiser_fast',duration=2.5,sr=22050*2,offset=0.5)
sample_rate = np.array(sample_rate)
mfccs = np.mean(librosa.feature.mfcc(y=X, sr=sample_rate, n_mfcc=13),axis=0)
featurelive = mfccs
livedf2 = featurelive
livedf2= pd.DataFrame(data=livedf2)
livedf2 = livedf2.stack().to_frame().T
twodim= np.expand_dims(livedf2, axis=2)
livepreds = loaded_model.predict(twodim, batch_size=32, verbose=1)

livepreds1=livepreds.argmax(axis=1)
liveabc = livepreds1.astype(int).flatten()
livepredictions = (lb.inverse_transform((liveabc)))
print(livepredictions)

[WenjingZhang0214]

You can find the emotion index list from here .You just need the variance liveabc, not the LabelEncoder instance lb.

[Aayush795]

@WenjingZhang0214 https://github.com/DimensionNXG/Speech-Emotion-Analyzer/blob/master/inference.py