Utilising Low Complexity CNNs to Lift Non-Local Redundancies in Video Coding.

Digital media is ubiquitous and produced in ever-growing quantities. This necessitates a constant evolution of compression techniques, especially for video, in order to maintain efficient storage and transmission. In this work, we aim at exploiting non-local redundancies in video data that remain difficult to erase for conventional video codecs We design convolutional neural networks with a particular emphasis on low memory and computational footprint. The parameters of those networks are trained on the fly, at encoding time, to predict the residual signal from the decoded video signal. After the training process has converged, the parameters are compressed and signalled as part of the code of the underlying video codec. The method can be applied to any existing video codec to increase coding gains while its low computational footprint allows for an application under resource-constrained conditions. Building on top of High Efficiency Video Coding, we achieve coding gains similar to those of pretrained denoising CNNs while only requiring about 1% of their computational complexity Through extensive experiments, we provide insights into the effectiveness of our network design decisions. In addition, we demonstrate that our algorithm delivers stable performance under conditions met in practical video compression: our algorithm performs without significant performance loss on very long random access segments (up to 256 frames) and with moderate performance drops can even be applied to single frames in high-resolution low delay settings.

Detection of acute periodontal pain from physiological signals.

OBJECTIVE: To investigate the feasibility of the detection of brief orofacial pain sensations from easily recordable physiological signals by means of machine learning techniques. APPROACH: A total of 47 subjects underwent periodontal probing and indicated each instance of pain perception by means of a push button. Simultaneously, physiological signals were recorded and, subsequently, autonomic indices were computed. By using the autonomic indices as input features of a classifier, a pain indicator based on fusion of the various autonomic mechanisms was achieved. Seven patients were randomly chosen for the test set. The rest of the data were utilized for the validation of several classifiers and feature combinations by applying leave-one-out-cross-validation. MAIN RESULTS: During the validation process the random forest classifier, using frequency spectral bins of the ECG, wavelet level energies of the ECG and PPG, PPG amplitude, and SPI as features, turned out to be the best pain detection algorithm. The final test of this algorithm on the independent test dataset yielded a sensitivity and specificity of 71% and 70%, respectively. SIGNIFICANCE: Based on these results, fusion of autonomic indices by applying machine learning techniques is a promising option for the detection of very brief instances of pain perception, that are not covered by the established indicators.