Incremental Knowledge Assimilation System (IKAS)

Mine countermeasures are an important aspect of Navy reconnaissance interests and reliable mine detection still remains a challenge. For automatic mine detection in underwater sonar applications, research has indicated that identifying the context of the scene will improve the recognition performance. Further, there is a need to constantly update and modify the representative knowledge base with information from new target and background samples. In Phase I, UtopiaCompression(UC) designed and demonstrated an Incremental Knowledge Assimilation System (IKAS) that identifies the environmental context in the scene and optimally incorporates new and relevant data samples into the knowledge base (classifier) in an incremental fashion with minimal memory and computation requirements. In Phase II, UC will be maturing this technology by incorporating active learning techniques that reduce the amount of operator input needed, improving statistical models for mine detection and classification, and enhancing the unsupervised context modeling module. The resulting system will be capable of detecting multiple targets in side-scan sonar images with varied seafloor backgrounds and classifying these targets as threats or harmless. IKAS’s ability to update its knowledge with minimal user input and detect mines reliably will make it a huge boon to mine countermeasures within the Navy, DoD, and beyond.