Examples of our expertise
Core AI – model generation
We can help you to protect core AI inventions, where the advances lie in the models, algorithms or mathematical methods that underpin the AI. Core AI inventions may be generally applicable or may be generated to solve a particular problem, such as how to implement AI on ordinary devices or how to train AI on lots of user-generated data without compromising user data privacy,
Core AI – dataset generation
Everyone knows the phrase “garbage in, garbage out”, i.e. that an AI model’s outputs are only as good as the data it was trained on. It can be difficult to obtain large and good quality datasets for use in training a model. Thus, synthetic dataset generation is an important part of AI model development and deployment. We can help you to protect inventions that relate to generating datasets for use in training AI models. Protecting the methods to generate the data, as well as the data itself may be important for your business, as there can be commercial value in both.
We have obtained granted patents for methods of generating datasets for shadow detection and removal, depth estimation, as well as tilt correction.
Applied AI – real-world applications of AI
We have drafted and prosecuted patent applications that utilise AI in a wide, and ever-expanding, variety of areas! Some examples:
We have also worked on inventions that span multiple disciplines, such as generating a biomarker signature for a particular cancer. For multi-disciplinary inventions, we pair-up with a colleague in the other discipline(s) to ensure that the patent application is prepared correctly, since the law on what can be patented and how it can be patented varies between the biotech and chemistry spaces.