Numerous imaging techniques are available for observing and interrogating biological samples, and several of them can be used consecutively to enable correlative analysis of different image modalities with varying resolutions and the inclusion of structural or molecular information. Achieving accurate registration of multimodal images is essential for the correlative analysis process, but it remains a challenging computer vision task with no widely accepted solution. Moreover, supervised registration methods require annotated data produced by experts, which is limited. To address this challenge, we propose a general unsupervised pipeline for multimodal image registration using deep learning. We provide a comprehensive evaluation of the proposed pipeline versus the current state-of-the-art image registration and style transfer methods on four types of biological problems utilizing different microscopy modalities. We found that style transfer of modality domains paired with fully unsupervised training leads to comparable image registration accuracy to supervised methods and, most importantly, does not require human intervention.

Surface Electromyography (sEMG) is to record muscles' electrical activity from a restricted area of the skin by using electrodes. The sEMG-based gesture recognition is extremely sensitive of inter-session and inter-subject variances. We propose a model and a deep-learning-based domain adaptation method to approximate the domain shift for recognition accuracy enhancement. Analysis performed on sparse and High-Density (HD) sEMG public datasets validate that our approach outperforms state-of-the-art methods.

This paper presents a machine learning approach to explore the phenetic relations of historical scripts and their glyphs. Its first step is the identification of the observable topological transformations in the development of the glyphs, and with the use of these transformations, the method collects the possible cognate glyphs by minimizing the necessary topological transformations between the glyphs. In these investigations, the phonetic properties of the graphemes were consistently considered. The second step of our method is selecting similarity groups of possible cognate glyphs by minimizing the differences of their topological properties. The third step is multidimensional scaling and different cluster analyses based on the similarity groups of the glyphs of the historical scripts in order to explore the phenetic relationships between these scripts. The resulting phenetic structure of the scripts could be used for paleographical research, especially in deciphering ancient hard-to-read inscriptions.

A method for optimization of MOS model parameters is introduced. The method is based upon dynamically fitting the measured frequency response of a reconfigurable ring oscillator. The ring oscillator contains two added switching transistors and two loading capacitors. By controlling the switching transistors, the capacitive load of the inverters and thus the oscillation frequency can be varied. The simulated and the measured frequencies are compared and the derived average error is minimized. Sequential iterations using combined mathematical methods are applied to the procedure. Simplifications are introduced in order to shorten the computational time. The effectiveness of the method on improving the accuracy of simulation is demonstrated with numerical examples.>