Efficient Facial Feature Learning with Wide Ensemble-based Convolutional Neural Networks

Abstract

Ensemble methods, traditionally built with independently trained de-correlated models, have proven to be efficient methods for reducing the remaining residual generalization error, which results in robust and accurate methods for realworld applications. In the context of deep learning, however, training an ensemble of deep networks is costly and generates high redundancy which is inefficient. In this paper, we present experiments on Ensembles with Shared Representations (ESRs) based on convolutional networks to demonstrate, quantitatively and qualitatively, their data processing efficiency and scalability to large-scale datasets of facial expressions. We show that redundancy and computational load can be dramatically reduced by varying the branching level of the ESR without loss of diversity and generalization power, which are both important for ensemble performance. Experiments on large-scale datasets suggest that ESRs reduce the remaining residual generalization error on the AffectNet and FER+ datasets, reach human-level performance, and outperform state-of-the-art methods on facial expression recognition in the wild using emotion and affect concepts.

The FaceChannel: A Fast & Furious Deep Neural Network for Facial Expression Recognition

“…We outperform their result by almost 6% as reported in Table 6. We also outperform the results reported in Miao et al [43], Li et al [44], and Siqueira et al [26] which employ different type of complex neural networks to learn facial expressions.…”

“…Most of these models, however, employ large and deep neural networks that demand a high computational power for training and re-adapting [23,24,10]. As a result, these models specialize in recognizing emotion expressions under conditions represented in the datasets they are trained with [25,26]. Thus, when these models are used to recognize facial expression under different conditions, not represented in the training data, they tend to perform poorly.…”

Across the Universe: Biasing Facial Representations Toward Non-Universal Emotions With the Face-STN

“…They focus on using a fine-tuned VGG13 encoder that updates all the convolutional layers. We also outperform the results Miao et al [60], Li et al [51], and Siqueira et al [72] reported, all of which employ different types of complex neural networks to learn facial expressions. On the FABO dataset, the Face-STN achieves higher results than reported in the literature, including Chen et al [18], who proposed a frame-based recognition and a bag-of-words-based model, or even Gunes et al [33] who used an SVM-based implementation.…”

“…Norm. [18] 66.5 % SHCNN [60] 86.54 % Bag of Words [18] 59.00 % TFE-JL [51] 84.30% SVM [33] 32.49 % ESR-9 [72] 87. Some of the models with which these datasets were evaluated seem to be outdated for other computer vision tasks.…”

Hey Human, If your Facial Emotions are Uncertain, You Should Use Bayesian Neural Networks!

“…Overall most methods for facial emotion recognition use classical neural networks, and Bayesian neural networks are not commonly used, even more recent work that uses ensembles like Siqueira et al [19] or Surace et al [20] do not consider the possibility of modeling output uncertainty, despite Lakshminarayanan et al [13] showing that ensembles are able to produce state of the art uncertainty quantification.…”