src package

Subpackages

• src.classification package
  • Subpackages
    • src.classification.fusion package
      • Submodules
      • src.classification.fusion.fusion_classifier module
      • Module contents
    • src.classification.image package
      • Submodules
      • src.classification.image.cross_attention_classifier module
      • src.classification.image.efficientnet_classifier module
      • src.classification.image.image_emotion_classifier module
      • src.classification.image.vgg16_classifier module
      • Module contents
    • src.classification.plant package
      • Submodules
      • src.classification.plant.dense_classifier module
      • src.classification.plant.lstm_classifier module
      • src.classification.plant.mfcc_cnn_classifier module
      • src.classification.plant.mfcc_resnet_classifier module
      • src.classification.plant.nn_classifier module
      • src.classification.plant.plant_emotion_classifier module
      • Module contents
    • src.classification.speech package
      • Submodules
      • src.classification.speech.byols_classifier module
      • src.classification.speech.gmm_classifier module
      • src.classification.speech.hmm_classifier module
      • src.classification.speech.hubert_classifier module
      • src.classification.speech.mfcc_lstm_classifier module
      • src.classification.speech.speech_emotion_classifier module
      • src.classification.speech.svm_classifier module
      • src.classification.speech.wav2vec2_classifier module
      • Module contents
    • src.classification.text package
      • Submodules
      • src.classification.text.bert_classifier module
      • src.classification.text.distilbert_classifier module
      • src.classification.text.nrclex_classifier module
      • src.classification.text.text_emotion_classifier module
      • Module contents
    • src.classification.watch package
      • Submodules
      • src.classification.watch.dense_classifier module
      • src.classification.watch.lstm_classifier module
      • src.classification.watch.nn_classifier module
      • src.classification.watch.random_forest_classifier module
      • src.classification.watch.transformer_classifier module
      • src.classification.watch.watch_emotion_classifier module
      • src.classification.watch.xgboost_classifier module
      • Module contents
  • Submodules
  • src.classification.classifier_factory module
  • src.classification.emotion_classifier module
  • Module contents
• src.data package
  • Submodules
  • src.data.balanced_image_data_reader module
  • src.data.balanced_plant_exp_reader module
  • src.data.balanced_watch_exp_reader module
  • src.data.classwise_speech_data_reader module
  • src.data.comparison_image_data_reader module
  • src.data.comparison_speech_data_reader module
  • src.data.comparison_text_data_reader module
  • src.data.data_factory module
  • src.data.data_reader module
  • src.data.experiment_data_reader module
  • src.data.fusion_data_reader module
  • src.data.image_data_reader module
  • src.data.plant_exp_reader module
  • src.data.speech_data_reader module
  • src.data.text_data_reader module
  • src.data.watch_exp_reader module
  • Module contents
• src.evaluation package
  • Submodules
  • src.evaluation.evaluator module
  • Module contents
• src.experiment package
  • Submodules
  • src.experiment.cv_experiment module
  • src.experiment.experiment module
  • Module contents
• src.utils package
  • Subpackages
    • src.utils.logging package
      • Submodules
      • src.utils.logging.base_logger module
      • src.utils.logging.pytorch_logger module
      • src.utils.logging.standard_logger module
      • src.utils.logging.tensorflow_logger module
      • Module contents
  • Submodules
  • src.utils.ground_truth module
  • src.utils.metrics module
  • src.utils.training module
  • Module contents

Submodules

src.emotion_set module

This file defines the emotion sets that are available in this work.

class src.emotion_set.AbstractEmotionSet(name: str, count: int, classes: Iterable[str])

Bases: object

This class implements the abstract interface for all emotion sets.

One problem in emotion detection is that a lot of different emotion sets are used throughout the past research. While one paper predicts only positive, neutral and negative emotions, another one might differentiate between 6 or more distinct emotions (joy, anger, fear, …). This class models this behaviour by defining different emotion sets.

get_emotions(indices: Union[int, ndarray]) → Union[str, ndarray]

This function returns the emotion strings for given indices

Parameters:

indices – The index or indices of the emotions

Returns:

Array of emotion strings or single emotion string

class src.emotion_set.EkmanEmotions

Bases: AbstractEmotionSet

Emotion set defined by Paul Ekman that is commonly used in research. It contains 6 basic emotions: anger, surprise, disgust, enjoyment, fear, and sadness

class src.emotion_set.EkmanNeutralEmotions

Bases: AbstractEmotionSet

Ekman Emotion set extended by a neutral state.

class src.emotion_set.EmotionMapper

Bases: object

map_emotion(emotion: str) → str

Maps an emotion to its equivalent in the NeutralEkman emotion set.

Parameters:

emotion – The name of the emotion to map to Ekman

Returns:

The name of the corresponding Ekman emotion

setup_map() → None

One time setup for the emotion map in the class.

class src.emotion_set.EmotionSetFactory

Bases: object

Factory class that generates emotion set instances

static generate(name: str) → AbstractEmotionSet

Method that creates and returns an instance of the emotion set specified by the name parameter

Parameters:

name – The name of the desired emotion set

Raises:

ValueError – Raised if the name does not represent an emotion set

Returns:

The created emotion set

class src.emotion_set.ThreeEmotions

Bases: AbstractEmotionSet

Simple emotion set with only three categories: positive, neutral, negative

Module contents

Package containing the entire source code