Data Augmentation for Audio-Visual Emotion Recognition with an Efficient Multimodal Conditional GAN

New Trends in Emotion Recognition Using Image Analysis by Neural Networks, A Systematic Review

Facial emotion recognition (FER) is a computer vision process aimed at detecting and classifying human emotional expressions. FER systems are currently used in a vast range of applications from areas such as education, healthcare, or public safety; therefore, detection and recognition accuracies are very important. Similar to any computer vision task based on image analyses, FER solutions are also suitable for integration with artificial intelligence solutions represented by different neural network varieties, especially deep neural networks that have shown great potential in the last years due to their feature extraction capabilities and computational efficiency over large datasets. In this context, this paper reviews the latest developments in the FER area, with a focus on recent neural network models that implement specific facial image analysis algorithms to detect and recognize facial emotions. This paper's scope is to present from historical and conceptual perspectives the evolution of the neural network architectures that proved significant results in the FER area. This paper endorses convolutional neural network (CNN)-based architectures against other neural network architectures, such as recurrent neural networks or generative adversarial networks, highlighting the key elements and performance of each architecture, and the advantages and limitations of the proposed models in the analyzed papers. Additionally, this paper presents the available datasets that are currently used for emotion recognition from facial expressions and micro-expressions. The usage of FER systems is also highlighted in various domains such as healthcare, education, security, or social IoT. Finally, open issues and future possible developments in the FER area are identified.

Generating synthetic personal health data using conditional generative adversarial networks combining with differential privacy

A large amount of personal health data that is highly valuable to the scientific community is still not accessible or requires a lengthy request process due to privacy concerns and legal restrictions. As a solution, synthetic data has been studied and proposed to be a promising alternative to this issue. However, generating realistic and privacy-preserving synthetic personal health data retains challenges such as simulating the characteristics of the patients' data that are in the minority classes, capturing the relations among variables in imbalanced data and transferring them to the synthetic data, and preserving individual patients' privacy. In this paper, we propose a differentially private conditional Generative Adversarial Network model (DP-CGANS) consisting of data transformation, sampling, conditioning, and network training to generate realistic and privacy-preserving personal data. Our model distinguishes categorical and continuous variables and transforms them into latent space separately for better training performance. We tackle the unique challenges of generating synthetic patient data due to the special data characteristics of personal health data. For example, patients with a certain disease are typically the minority in the dataset and the relations among variables are crucial to be observed. Our model is structured with a conditional vector as an additional input to present the minority class in the imbalanced data and maximally capture the dependency between variables. Moreover, we inject statistical noise into the gradients in the networking training process of DP-CGANS to provide a differential privacy guarantee. We extensively evaluate our model with state-of-the-art generative models on personal socio-economic datasets and real-world personal health datasets in terms of statistical similarity, machine learning performance, and privacy measurement. We demonstrate that our model outperforms other comparable models, especially in capturing the dependence between variables. Finally, we present the balance between data utility and privacy in synthetic data generation considering the different data structures and characteristics of real-world personal health data such as imbalanced classes, abnormal distributions, and data sparsity.

The multi-level classification network (MCN) with modified residual U-Net for ischemic stroke lesions segmentation from ATLAS

Ischemic and hemorrhagic strokes are two major types of internal brain injury. 3D brain MRI is suggested by neurologists to examine the brain. Manual examination of brain MRI is very sensitive and time-consuming task. However, automatic diagnosis can assist doctors in this regard. Anatomical Tracings of Lesions After Stroke (ATLAS) is publicly available dataset for research experiments. One of the major issues in medical imaging is class imbalance. Similarly, pixel representation of stroke lesion is less than 1% in ATLAS. Second major challenge in this dataset is inter-class similarity. A multi-level classification network (MCN) is proposed for segmentation of ischemic stroke lesions. MCN consists of three cascaded discrete networks. The first network designed to reduce the slice level class imbalance, where a classifier model is trained to extract the slices of stroke lesions from a whole brain MRI volume. The interclass similarity cause to produce false positives in segmented output. Therefore, all extracted stroke slices were divided into overlapping patches (64 × 64) and carried to the second network. The task associated with second network is to classify the patches comprises of stroke lesion. The third network is a 2D modified residual U-Net that segments out the stroke lesions from the patches extracted by the second network. MCN achieved 0.754 mean dice score on test dataset which is higher than the other state-of-the-art methods on the same dataset.

Affective video recommender systems: A survey

Traditional video recommendation provides the viewers with customized media content according to their historical records (e.g., ratings, reviews). However, such systems tend to generate terrible results if the data is insufficient, which leads to a cold-start problem. An affective video recommender system (AVRS) is a multidiscipline and multimodal human-robot interaction (HRI) system, and it incorporates physical, physiological, neuroscience, and computer science subjects and multimedia resources, including text, audio, and video. As a promising research domain, AVRS employs advanced affective analysis technologies in video resources; therefore, it can solve the cold-start problem. In AVRS, the viewers’ emotional responses can be obtained from various techniques, including physical signals (e.g., facial expression, gestures, and speech) and internal signals (e.g., physiological signals). The changes in these signals can be detected when the viewers face specific situations. The physiological signals are a response to central and autonomic nervous systems and are mostly involuntarily activated, which cannot be easily controlled. Therefore, it is suitable for reliable emotion analysis. The physical signals can be recorded by a webcam or recorder. In contrast, the physiological signals can be collected by various equipment, e.g., psychophysiological heart rate (HR) signals calculated by echocardiogram (ECG), electro-dermal activity (EDA), and brain activity (GA) from electroencephalography (EEG) signals, skin conductance response (SCR) by a galvanic skin response (GSR), and photoplethysmography (PPG) estimating users’ pulse. This survey aims to provide a comprehensive overview of the AVRS domain. To analyze the recent efforts in the field of affective video recommendation, we collected 92 relevant published articles from Google Scholar and summarized the articles and their key findings. In this survey, we feature these articles concerning AVRS from different perspectives, including various traditional recommendation algorithms and advanced deep learning-based algorithms, the commonly used affective video recommendation databases, audience response categories, and evaluation methods. Finally, we conclude the challenge of AVRS and provide the potential future research directions.

A Preliminary Investigation on Frequency Dependant Cues for Human Emotions

The recent advances in Human-Computer Interaction and Artificial Intelligence have significantly increased the importance of identifying human emotions from different sensory cues. Hence, understanding the underlying relationships between emotions and sensory cues have become a subject of study in many fields including Acoustics, Psychology, Psychiatry, Neuroscience and Biochemistry. This work is a preliminary step towards investigating cues for human emotion on a fundamental level by aiming to establish relationships between tonal frequencies of sound and emotions. For that, an online perception test is conducted, in which participants are asked to rate the perceived emotions corresponding to each tone. The results show that a crossover point for four primary emotions lies in the frequency range of 417–440 Hz, thus consolidating the hypothesis that the frequency range of 432–440 Hz is neutral from human emotion perspective. It is also observed that the frequency dependant relationships between emotion pairs Happy—Sad, and Anger—Calm are approximately mirrored symmetric in nature.

Multi-Attention Module for Dynamic Facial Emotion Recognition

Video-based dynamic facial emotion recognition (FER) is a challenging task, as one must capture and distinguish tiny facial movements representing emotional changes while ignoring the facial differences of different objects. Recent state-of-the-art studies have usually adopted more complex methods to solve this task, such as large-scale deep learning models or multimodal analysis with reference to multiple sub-models. According to the characteristics of the FER task and the shortcomings of existing methods, in this paper we propose a lightweight method and design three attention modules that can be flexibly inserted into the backbone network. The key information for the three dimensions of space, channel, and time is extracted by means of convolution layer, pooling layer, multi-layer perception (MLP), and other approaches, and attention weights are generated. By sharing parameters at the same level, the three modules do not add too many network parameters while enhancing the focus on specific areas of the face, effective feature information of static images, and key frames. The experimental results on CK+ and eNTERFACE’05 datasets show that this method can achieve higher accuracy.