预测结果 - TensorFlow

Question

除了模型的准确性之外，我还尝试打印预测结果和标签。 我不确定我在这里做错了什么

for mfcc, label in test_data:
 prediction = tflite_inference(mfcc, tflite_path)
 predicted_indices.append(np.squeeze(tf.argmax(prediction, axis=1)))

strlabel="C:/tmp/speech_commands_train/conv_labels.txt"
labels_list= [line.rstrip() for line in tf.io.gfile.GFile(strlabel)]

top_k = prediction.argsort()[-5:][::-1]


for node_id in top_k:
 human_string = labels_list[node_id]
 score = predicted_indices[node_id]
print('%s (score = %.5f)' % (human_string, score))

test_accuracy = calculate_accuracy(predicted_indices, expected_indices)
confusion_matrix = tf.math.confusion_matrix(expected_indices, predicted_indices,
                                        num_classes=model_settings['label_count'])

` 错误信息

human_string = labels_list[node_id] TypeError: only integer scalar arrays can be converted to a scalar index

提前感谢您的帮助。

Answer 1

编辑的答案（在对问题进行了一些澄清之后）： 在这里，我假设 prediction 变量是您的模型对于单个输入的输出。在此假设下，您的 top_k 应包含概率最高的前 5 个索引。 为此，您应该执行以下操作：

1. 重塑您的 predictions 变量：

predictions = predictions.reshape(-1) # this will make the predicitions a vector

1. 获取top_k

# this step is same but this time the output will be a vector instead of a matrix
top_k = prediction.argsort()[-5:][::-1] 

1. 使用循环

# This is also same but as the `top_k` is a vector instead of a matrix there 
# won't be any issues/errors.
for node_id in top_k:
 human_string = labels_list[node_id]
 score = predicted_indices[node_id]
 print('%s (score = %.5f)' % (human_string, score))