This inventor holds 1 USPTO granted patent and 1 published patent application. Top assignee: Royal Bank of Canada. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2026
Title: Maryna Karpusha: Innovator in Artificial Neural Networks
Introduction
Maryna Karpusha is a prominent inventor based in Vancouver, Canada. She has made significant contributions to the field of artificial intelligence, particularly in the development of multi-modal artificial neural networks. Her innovative work has the potential to enhance the way machines process and analyze complex data.
Latest Patents
Maryna holds a patent for a "Multi-modal artificial neural network and a self-supervised learning method for training same." This invention involves a sophisticated learning method that processes augmented views of input data through simple Siamese networks. The method determines cross-modal losses and total losses to train the networks effectively. The trained network can analyze multi-modal content, such as video with audio tracks, and can also train a Multi-Modal Multi-Head Network (M3HN) for processing various representations.
Career Highlights
Maryna Karpusha is currently employed at the Royal Bank of Canada, where she applies her expertise in artificial intelligence to drive innovation. Her work focuses on developing advanced neural network models that can improve data analysis and machine learning applications.
Collaborations
Maryna collaborates with talented individuals such as Sri Harsha Dumpala and Ainaz Hajimoradlou. These partnerships enhance her research and development efforts, contributing to the advancement of technology in her field.
Conclusion
Maryna Karpusha is a trailblazer in the realm of artificial intelligence, with her patent and work at the Royal Bank of Canada showcasing her innovative spirit. Her contributions are paving the way for future advancements in multi-modal neural networks.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
