Sonasoft Corp. (OTCQB: SSFT) has filed an important new patent application in a key area of AI. This invention solves one of the hardest problems in AI. Namely, how to cope when your training data is noisy, limited, or unpredictable. This approach lies at the heart of NuGene, Sonasoft’s AI bot engine, which helps companies put AI at the heart of their business processes.
Sonasoft’s invention relates to convolutional neural networks or CNNs. CNNs are the artificial brains that drive computer vision systems. They also power deep learning, where computers are able to learn to outperform humans. CNNs are based on the structures found in the human brain. Like a human child, a CNN has to learn from the data provided to it. Unfortunately, they only work well when there are large amounts of consistent training data. However, if that training data is sparse or noisy, they stop performing.
The invention, titled “Convolutional Hierarchical Temporal Memory Modules”, addresses the problem in a novel way. It provides a robust and noise-tolerant system that will create reliable models using far less training data than traditional CNN-based systems. Moreover, the resulting models are smaller, allowing them to run on less complex hardware. “This has the potential to allow the development of classification or predictive models with limited or lower quality data,” said Max Lee, Head of Engineering.
Like CNNs themselves, the approach is inspired by biology, as well as systems like Numenta’s HTM and Applied Brain Research’s spiking neural network. As Max puts it, “By studying how intelligence works in biological systems, we can continue to improve and provide more robust artificial intelligence. AI has always learned from nature and this takes us another step towards true artificial general intelligence.”
Mike Khanna, Sonasoft’s CEO, explains, “This new technique will allow more companies to benefit from the disruptive impact of AI. Sonasoft NuGene already incorporates the technology, which allows us to create AI models even when we only get limited or noisy data. This will help our customers leverage AI throughout their business. By patenting the technology, we defend our shareholders’ and investors’ interests while ensuring the approach is shared with the wider research community.”
Mike Khanna, CEO Sonasoft Corporation. Phone: (408) 708-4000 X7104
This release contains statements that constitute forward-looking statements. These statements appear in a number of places in this release and include all statements that are not statements of historical fact regarding the intent, belief or current expectations of the Company, its directors or its officers with respect to, among other things: (i) the Company’s financing plans; (ii) trends affecting the Company’s financial condition or results of operations; (iii) the Company’s growth strategy and operating strategy; and (iv) the declaration and payment of dividends. The words “may,” “would,” “will,” “expect,” “estimate,” “anticipate,” “believe,” “intend,” and similar expressions and variations thereof are intended to identify forward-looking statements. Investors are cautioned that any such forward-looking statements are not guarantees of future performance and involve risks and uncertainties, many of which are beyond the Company’s ability to control and that actual results may differ materially from those projected in the forward-looking statements as a result of various factors.