A Proposal for Multimodal Emotion Recognition Using Aural Transformers and Action Units on RAVDESS Dataset

Emotion recognition is attracting the attention of the research community due to its multiple applications in different fields, such as medicine or autonomous driving. In this paper, we proposed an automatic emotion recognizer system that consisted of a speech emotion recognizer (SER) and a facial emotion recognizer (FER). For the SER, we evaluated a pre-trained xlsr-Wav2Vec2.0 transformer using two transfer-learning techniques: embedding extraction and fine-tuning. The best accuracy results were achieved when we fine-tuned the whole model by appending a multilayer perceptron on top of it, confirming that the training was more robust when it did not start from scratch and the previous knowledge of the network was similar to the task to adapt. Regarding the facial emotion recognizer, we extracted the Action Units of the videos and compared the performance between employing static models against sequential models. Results showed that sequential models beat static models by a narrow difference. Error analysis reported that the visual systems could improve with a detector of high-emotional load frames, which opened a new line of research to discover new ways to learn from videos. Finally, combining these two modalities with a late fusion strategy, we achieved 86.70% accuracy on the RAVDESS dataset on a subject-wise 5-CV evaluation, classifying eight emotions. Results demonstrated that these modalities carried relevant information to detect users’ emotional state and their combination allowed to improve the final system performance.

OPTIMIZED ACTION UNITS FEATURES FOR EFFICIENT DESIGN OF DECEPTION DETECTION SYSTEM

Deception detection is becoming an interesting filed in different areas related to security, criminal investigation, law enforcement and terrorism detection. Recently non-verbal features have become essential features for deception detection process. One of the most important kind of these features is facial expression. The importance of these expressions come from the idea that Human face contain different expressions each of which is directly related to a certain state. In this research paper, facial expressions' data are collected for 102 participants (25 women and 77 men) as video clips. There are 504 clips for lie response and 384 for truth response (total 888 video clips). Facial expressions in a form of Action Units (AUs) are extracted for each frame with video clip. The AUs are encoded based on Facial Action Coding System (FACS) which are 18 AUs. These are: AU 1, 2, 4, 5, 6, 7, 9, 10, 12, 14, 15, 17, 20, 23, 25, 26, 28 and 45. Based on the collected data, only six AUs are the most effective and have a direct impact on the discrimination process between liar and truth teller. These AUs are AU 6, 7, 10, 12, 14 and 28

Classify Elderly Pain Severity from Open-Source Automatically Video Clip Facial Action Units Analysis. A Study from the Integrated Pain Artificial Intelligence Network (I-PAIN) Data Repository.

Background: It is increasingly interesting to monitor pain severity in elderly individuals by applying machine learning models. In previous studies, OpenFace© - a well-known automated facial analysis algorithm, was used to detect facial action units (FAUs) that initially need long hours of human coding. However, OpenFace© developed from the dataset that dominant young Caucasians who were illicit pain in the lab. Therefore, this study aims to evaluate the accuracy and feasibility of the model using data from OpenFace© to classify pain severity in elderly Asian patients in clinical settings.Methods: Data from 255 Thai individuals with chronic pain were collected at Chiang Mai Medical School Hospital. The phone camera recorded faces for 10 seconds at a 1-meter distance briefly after the patients provided self-rating pain severity. For those unable to self-rate, the video was recorded just after the move, which illicit pain. The trained assistant rated each video clip for the Pain Assessment in Advanced Dementia (PAINAD). The classification of pain severity was mild, moderate, or severe. OpenFace© process video clip into 18 FAUs. Five classification models were used, including logistic regression, multilayer perception, naïve Bayes, decision tree, k-nearest neighbors (KNN), and support vector machine (SVM). Results: Among the models that included only FAU described in the literature (FAUs 4, 6, 7, 9, 10, 25, 26, 27 and 45), multilayer perception yielded the highest accuracy of 50%. Among the machine learning selection features, the SVM model for FAU 1, 2, 4, 7, 9, 10, 12, 20, 25, 45, and gender yielded the best accuracy of 58%. Conclusion: Our open-source automatic video clip facial action unit analysis experiment was not robust for classifying elderly pain. Retraining facial action unit detection algorithms, enhancing frame selection strategies, and adding pain-related functions may improve the accuracy and feasibility of the model.

Synthesising Facial Macro- and Micro-Expressions Using Reference Guided Style Transfer

Long video datasets of facial macro- and micro-expressions remains in strong demand with the current dominance of data-hungry deep learning methods. There are limited methods of generating long videos which contain micro-expressions. Moreover, there is a lack of performance metrics to quantify the generated data. To address the research gaps, we introduce a new approach to generate synthetic long videos and recommend assessment methods to inspect dataset quality. For synthetic long video generation, we use the state-of-the-art generative adversarial network style transfer method—StarGANv2. Using StarGANv2 pre-trained on the CelebA dataset, we transfer the style of a reference image from SAMM long videos (a facial micro- and macro-expression long video dataset) onto a source image of the FFHQ dataset to generate a synthetic dataset (SAMM-SYNTH). We evaluate SAMM-SYNTH by conducting an analysis based on the facial action units detected by OpenFace. For quantitative measurement, our findings show high correlation on two Action Units (AUs), i.e., AU12 and AU6, of the original and synthetic data with a Pearson’s correlation of 0.74 and 0.72, respectively. This is further supported by evaluation method proposed by OpenFace on those AUs, which also have high scores of 0.85 and 0.59. Additionally, optical flow is used to visually compare the original facial movements and the transferred facial movements. With this article, we publish our dataset to enable future research and to increase the data pool of micro-expressions research, especially in the spotting task.

Effect of transportation and social isolation on facial expressions of healthy horses

Horses have the ability to generate a remarkable repertoire of facial expressions, some of which have been linked to the affective component of pain. This study describes the facial expressions in healthy horses free of pain before and during transportation and social isolation, which are putatively stressful but ordinary management procedures. Transportation was performed in 28 horses by subjecting them to short-term road transport in a horse trailer. A subgroup (n = 10) of these horses was also subjected to short-term social isolation. During all procedures, a body-mounted, remote-controlled heart rate monitor provided continuous heart rate measurements. The horses’ heads were video-recorded during the interventions. An exhaustive dataset was generated from the selected video clips of all possible facial action units and action descriptors, time of emergency, duration, and frequency according to the Equine Facial Action Coding System (EquiFACS). Heart rate increased during both interventions (p<0.01), confirming that they caused disruption in sympato-vagal balance. Using the current method for ascribing certain action units (AUs) to specific emotional states in humans and a novel data-driven co-occurrence method, the following facial traits were observed during both interventions: eye white increase (p<0.001), nostril dilator (p<0.001), upper eyelid raiser (p<0.001), inner brow raiser (p = 0.042), tongue show (p<0.001). Increases in ‘ear flicker’ (p<0.001) and blink frequency (p<0.001) were also seen. These facial actions were used to train a machine-learning classifier to discriminate between the high-arousal interventions and calm horses, which achieved at most 79% accuracy. Most facial features identified correspond well with previous findings on behaviors of stressed horses, for example flared nostrils, repetitive mouth behaviors, increased eye white, tongue show, and ear movements. Several features identified in this study of pain-free horses, such as dilated nostrils, eye white increase, and inner brow raiser, are used as indicators of pain in some face-based pain assessment tools. In order to increase performance parameters in pain assessment tools, the relations between facial expressions of stress and pain should be studied further.