We empower your edtech with AI

In his research paper Development of Supervised Speaker Diarization System Based on the PyAnnote Audio Processing Library, Yuriy challenges various approaches in speaker diarization and the finds the architecture that delivers the most accurate results.

Dmytro compared deep speaker embedding models—WavLM, TitaNet, ECAPA, PyAnnote—on a dataset of short, non-English clips, revealing TitaNet and ECAPA as superior for speaker verification with the lowest Equal Error Rates.

Michael aimed to enhance humanoid robots, like Pepper, to handle multi-talker interactions beyond the traditional Read Eval Print Loop (REPL) limitations. By applying smart beamforming and sensor fusion, his research showed significant improvements in the robot's ability to engage with several speakers at once, marking a leap forward in robotic communication capabilities.

This paper we propose a concept of adaptive workflow system for scientific project collaboration. The process is so unique it became a part of the Communications in Computer and Information Science book, published in 2023 by Springer.

In Vitalii’s paper “Application of Deep Neural Networks for EEG Signal Processing in Brain-controlled Wheeled Robotic Platform”, the main task was to create and research the possibility of application of the deep learning technologies to classification of the filtered signals (frequency band of the EEG Alpha-waves) under relatively low data volume scenario.