Conversations between two peers exhibit a large amount of emotional content that dynamically creates an emotional climate (EC) during the conversation. The recognition of this EC using artificial intelligence (AI), gives an idea of how the conversation is emotionally interpreted by both interlocutors and external parties. This paper presents a new method for EC detection called DeepBispec that is based on deep bispectral processing of conversational speech. The latter is segmented based on emotional labels and subjected to windowed bispectral analysis. The calculated 2D-bispectrums are inputted as colored images to a convolutional neural network (CNN). The latter detects and extracts features from the bispectrum images, that are then fused with affect dynamics (AD) to classify arousal (A) and valence (V) into (low/high) classes. Extensive experiments on the IEMOCAP dataset with 2D emotions (i.e., A and V) show that DeepBispec outperforms previous state-of-the-art methods, achieving an accuracy of 0.826A/0.749V, sensitivity of 0.898A/0.774V, and area under the curve (AUC) of 0.845A/0.824V. The findings reveal the effectiveness of DeepBispec in detecting the emotional tone of conversations between peers, providing a deeper understanding of the emotional dynamics at play in social interactions.
Gait, or the unique way every human walks is widely recognised as an important indicator of overall health and predictor of fall risks. Recently, machine learning based classifiers have been acknowledged and adopted to monitor personal safety specially to detect or predict falls as health and safety anomalies among the older population. It is well known that other animals and birds, including chickens can also suffer from health problems that frequently manifest themselves as gait disorders, one of which is known as lameness. Comparing broiler lameness detection with human fall prediction and detection is meaningful for many reasons, one possible argument being that there are very few bipedal animal models, of which poultry is one. In our multidisciplinary collaborative search for a possible AI-based solution to detecting early lameness among broiler chickens, we discovered formidable challenges in applying anomaly detection techniques. However, prior to this work we had already gained valuable experience in gait anomaly detection for humans. This extension of our experience from humans to broilers has revealed new challenges in anomaly detection that are also relevant to human gait analysis, particularly in monitoring gait to enhance fall prediction and detection through assistive technologies.
Predicting and understanding the changes in cognitive performance, especially after a longitudinal intervention, is a fundamental goal in neuroscience. Longitudinal brain stimulation-based interventions like transcranial direct current stimulation (tDCS) induce short-term changes in the resting membrane potential and influence cognitive processes. However, very little research has been conducted on predicting these changes in cognitive performance post-intervention. In this research, we intend to address this gap in the literature by employing different EEG-based functional connectivity analyses and machine learning algorithms to predict changes in cognitive performance in a complex multitasking task. Forty subjects were divided into experimental and active-control conditions. On Day 1, all subjects executed a multitasking task with simultaneous 32-channel EEG being acquired. From Day 2 to Day 7, subjects in the experimental condition undertook 15 minutes of 2mA anodal tDCS stimulation during task training. Subjects in the active-control condition undertook 15 minutes of sham stimulation during task training. On Day 10, all subjects again executed the multitasking task with EEG acquisition. Source-level functional connectivity metrics, namely phase lag index and directed transfer function, were extracted from the EEG data on Day 1 and Day 10. Various machine learning models were employed to predict changes in cognitive performance. Results revealed that the multi-layer perceptron and directed transfer function recorded a cross-validation training RMSE of 5.11% and a test RMSE of 4.97%. We discuss the implications of our results in developing real-time cognitive state assessors for accurately predicting cognitive performance in dynamic and complex tasks post-tDCS intervention
A methodology based on deep recurrent models for maritime surveillance, over publicly available Automatic Identification System (AIS) data, is presented in this paper. The setup employs a deep Recurrent Neural Network (RNN)-based model, for encoding and reconstructing the observed ships’ motion patterns. Our approach is based on a thresholding mechanism, over the calculated errors between observed and reconstructed motion patterns of maritime vessels. Specifically, a deep-learning framework, i.e. an encoder-decoder architecture, is trained using the observed motion patterns, enabling the models to learn and predict the expected trajectory, which will be compared to the effective ones. Our models, particularly the bidirectional GRU with recurrent dropouts, showcased superior performance in capturing the temporal dynamics of maritime data, illustrating the potential of deep learning to enhance maritime surveillance capabilities. Our work lays a solid foundation for future research in this domain, highlighting a path toward improved maritime safety through the innovative application of technology.
The environmental hazards and climate change effects causes serious problems in land and coastal areas. A solution to this problem can be the periodic monitoring over critical areas, like coastal region with heavy industrial activity (i.e., ship-buildings) or areas where a disaster (i.e., oil-spill) has occurred. Today there are several Earth and non-Earth Observation data available from several data providers. These data are huge in size and usually it is needed to combine several data from multiple sources (i.e., data with format differences) for a more effective evaluation. For addressing these issues, this work proposes the Ocean-DC framework as a solution in data harmonization and homogenization. A strong advantage of this Data Cube implementation is the generation of a single NetCDF product that contains Earth Observation data of several data types (i.e., Landsat-8 and Sentinel-2). To evaluate the effectiveness and efficiency of the Ocean-DC implementation, it is examined a case study of an oil-spill in Saronic gulf in September of 2017. The generated 4D Data Cube considers both Landsat-8,9 and Sentinel-2 products for a time-series analysis, before, during, and after the oil-spill event. The Ocean-DC framework successfully generated a NetCDF product, containing all the necessary remote sensing products for monitoring the oil-spill disaster in the Saronic gulf.
