Deep Convolutional Neural Networks (CNNs) boosted the artificial intelligence revolution, especially after 2015. Although CNNs provide impressive performance for classification and relational reasoning on data patterns, they also require application-oriented structural optimization.This motivates the developers and researchers to work on morphologically adaptive CNNs where the network topology is evolved as processed data and experience of the system increase. On the other hand, this also increases the level of time and space complexity. In other words, GPU/CPU processing power and memory requirement exponentially raise. Neuromorphic computing intends to provide novel solutions for efficient utilization of hardware resources against increasing complexity for implementation of morphologically adaptive machine learning systems. For this reason, neuromorphic computing has become a forthcoming milestone as the next generation of AI after the development of deep convolutional networks.
An important target of the neuromorphic systems is high-level cognitive functions that mimic human intelligence and provides situational awareness. In order to reduce power consumption with efficient information coding, one way of neuromorphic computing implementation is using dedicated hardware that depends on parallelism instead of sequential logical processing solutions.
Computation based on conventional digital computer architectures and materials is approaching to physical limitations. Due to the development of AI applications and their rising economic impact, nature-inspired emulation of cognitive functions has become a major target within new computational paradigms. Neuromorphic computing, also known as, Neuromorphic engineering is a concept describing the use of integrated systems containing morphologically adaptive analog, analog & digital hybrid circuits or equivalent circuit simulations to mimic neuro-biological architectures present in the nervous system.
Situational awareness and context awareness are becoming key factors in cybersecurity because of exponentially increasing amount of threats both in terms of variety and their complexity. Monitoring, diagnostics, defense, and recovery requires a high level of cognition in order to sustain security in cyberspace. For this reason, CRYPTTECH AI team works on the development of high level of cognition depending on situational awareness on morphologically adaptive platforms for cybersecurity. We work on state of the art implementation of morphologically adaptive data coding and machine learning solutions on conventional computational systems as well as neuromorphic computing with dedicated hardware for high-performance cybersecurity systems.
Computation based on conventional digital computer architectures and materials is approaching to physical limitations. Due to the development of AI applications and their rising economic impact, nature-inspired emulation of cognitive functions has become a major target within new computational paradigms. Neuromorphic computing, also known as, Neuromorphic engineering is a concept describing the use of integrated systems containing morphologically adaptive analog, analog & digital hybrid circuits or equivalent circuit simulations to mimic neuro-biological architectures present in the nervous system.
Situational awareness and context awareness are becoming key factors in cybersecurity because of exponentially increasing amount of threats both in terms of variety and their complexity. Monitoring, diagnostics, defense, and recovery requires a high level of cognition in order to sustain security in cyberspace. For this reason, CRYPTTECH AI team works on the development of high level of cognition depending on situational awareness on morphologically adaptive platforms for cybersecurity. We work on state of the art implementation of morphologically adaptive data coding and machine learning solutions on conventional computational systems as well as neuromorphic computing with dedicated hardware for high-performance cybersecurity systems.