Abstract: Lifelong Deep Neural Network (L-DNN) technology revolutionizes Deep Learning by enabling fast, post-deployment learning without extensive training, heavy computing resources, or massive data storage. It uses a representation-rich, DNN-based subsystem (Module A) with a fast-learning subsystem (Module B) to learn new features quickly without forgetting previously learned features. Compared to a conventional DNN, L-DNN uses much less data to build robust networks, dramatically shorter training time, and learning on-device instead of on servers. It can add new knowledge without re-training or storing data. As a result, an edge device with L-DNN can learn continuously after deployment, eliminating massive costs in data collection and annotation, memory and data storage, and compute power. This fast, local, on-device learning can be used for security, supply chain monitoring, disaster and emergency response, and drone-based inspection of infrastructure and properties, among other applications.

Abstract: Industrial quality control is challenging for artificial neural networks (ANNs) and deep neural networks (DNNs) because of the nature of the processed data: there is an abundance of consistent data representing good products, but little data representing bad products. In quality control, the task is changed from conventional DNN task of “recognize what I learned best” to “recognize what I have never seen before.” Lifelong DNN (L-DNN) technology is a hybrid semi-supervised neural architecture that combines the ability of DNNs to be trained, with high precision, on known classes, while being sensitive to any number of unknown classes or class variations. When used for industrial inspection, L-DNN exploits its ability to learn with little and highly unbalanced data. L-DNN's real-time learning capability takes advantage of rare cases of poor-quality products that L-DNN encounters after deployment. L-DNN can be applied to industrial inspections and manufacturing quality control.

Abstract: Today, artificial neural networks are trained on large sets of manually tagged images. Generally, for better training, the training data should be as large as possible. Unfortunately, manually tagging images is time consuming and susceptible to error, making it difficult to produce the large sets of tagged data used to train artificial neural networks. To address this problem, the inventors have developed a smart tagging utility that uses a feature extraction unit and a fast-learning classifier to learn tags and tag images automatically, reducing the time to tag large sets of data. The feature extraction unit and fast-learning classifiers can be implemented as artificial neural networks that associate a label with features extracted from an image and tag similar features from the image or other images with the same label. Moreover, the smart tagging system can learn from user adjustment to its proposed tagging. This reduces tagging time and errors.

Abstract: Today, artificial neural networks are trained on large sets of manually tagged images. Generally, for better training, the training data should be as large as possible. Unfortunately, manually tagging images is time consuming and susceptible to error, making it difficult to produce the large sets of tagged data used to train artificial neural networks. To address this problem, the inventors have developed a smart tagging utility that uses a feature extraction unit and a fast-learning classifier to learn tags and tag images automatically, reducing the time to tag large sets of data. The feature extraction unit and fast-learning classifiers can be implemented as artificial neural networks that associate a label with features extracted from an image and tag similar features from the image or other images with the same label. Moreover, the smart tagging system can learn from user adjustment to its proposed tagging. This reduces tagging time and errors.

Abstract: The system and methods disclosed herein include a runtime architecture that takes a nonspecific set of systems of differential equations, distributes them across the network, and iteratively integrates them through time with a possibility to output the results on every iteration. Embodiments of the disclosed system may be used for neural computation or any other suitable application. Embodiments can be used as a standalone engine or as part of another computational system for massively parallel numerical integration of a data-driven dynamical system.

Abstract: The systems and methods disclosed herein include a path integration system that calculates optic flow, infers angular velocity from the flow field, and incorporates this velocity estimate into heading calculations. The resulting system fuses heading estimates from accelerometers, 5 gyroscopes, engine torques, and optic flow to determine self-localization. The system also includes a motivational system that implements a reward drive, both positive and negative, into the system. In some implementations, the drives can include: a) a curiosity drive that encourages exploration of new areas, b) a resource drive that attracts the agent towards the recharging base when the battery is low, and c) a mineral reward drive that attracts the agent 10 towards previously explored scientific targets.

Abstract: Sensory processing of visual, auditory, and other sensor information (e.g., visual imagery, LIDAR, RADAR) is conventionally based on “stovepiped,” or isolated processing, with little interactions between modules. Biological systems, on the other hand, fuse multi-sensory information to identify nearby objects of interest more quickly, more efficiently, and with higher signal-to-noise ratios. Similarly, examples of the OpenSense technology disclosed herein use neurally inspired processing to identify and locate objects in a robot's environment. This enables the robot to navigate its environment more quickly and with lower computational and power requirements.

Abstract: Sensory processing of visual, auditory, and other sensor information (e.g., visual imagery, LIDAR, RADAR) is conventionally based on “stovepiped,” or isolated processing, with little interactions between modules. Biological systems, on the other hand, fuse multi-sensory information to identify nearby objects of interest more quickly, more efficiently, and with higher signal-to-noise ratios. Similarly, examples of the OpenSense technology disclosed herein use neurally inspired processing to identify and locate objects in a robot's environment. This enables the robot to navigate its environment more quickly and with lower computational and power requirements.

Abstract: The system and methods disclosed herein include a runtime architecture that takes a nonspecific set of systems of differential equations, distributes them across the network, and iteratively integrates them through time with a possibility to output the results on every iteration. Embodiments of the disclosed system may be used for neural computation or any other suitable application. Embodiments can be used as a standalone engine or as part of another computational system for massively parallel numerical integration of a data-driven dynamical system.

Abstract: The systems and methods disclosed herein include a path integration system that calculates optic flow, infers angular velocity from the flow field, and incorporates this velocity estimate into heading calculations. The resulting system fuses heading estimates from accelerometers, gyroscopes, engine torques, and optic flow to determine self-localization. The system also includes a motivational system that implements a reward drive, both positive and negative, into the system. In some implementations, the drives can include: a) a curiosity drive that encourages exploration of new areas, b) a resource drive that attracts the agent towards the recharging base when the battery is low, and c) a mineral reward drive that attracts the agent towards previously explored scientific targets.

Abstract: Sensory processing of visual, auditory, and other sensor information (e.g., visual imagery, LIDAR, RADAR) is conventionally based on “stovepiped,” or isolated processing, with little interactions between modules. Biological systems, on the other hand, fuse multi-sensory information to identify nearby objects of interest more quickly, more efficiently, and with higher signal-to-noise ratios. Similarly, examples of the OpenSense technology disclosed herein use neurally inspired processing to identify and locate objects in a robot's environment. This enables the robot to navigate its environment more quickly and with lower computational and power requirements.

Abstract: Sensory processing of visual, auditory, and other sensor information (e.g., visual imagery, LIDAR, RADAR) is conventionally based on “stovepiped,” or isolated processing, with little interactions between modules. Biological systems, on the other hand, fuse multi-sensory information to identify nearby objects of interest more quickly, more efficiently, and with higher signal-to-noise ratios. Similarly, examples of the OpenSense technology disclosed herein use neurally inspired processing to identify and locate objects in a robot's environment. This enables the robot to navigate its environment more quickly and with lower computational and power requirements.

Abstract: An accelerator system is implemented on an expansion card comprising a printed circuit board having (a) one or more graphics processing units (GPUs), (b) two or more associated memory banks (logically or physically partitioned), (c) a specialized controller, and (d) a local bus providing signal coupling compatible with the PCI industry standards. The controller handles most of the primitive operations to set up and control GPU computation. Thus, the computer's central processing unit (CPU) can be dedicated to other tasks. In this case a few controls (simulation start and stop signals from the CPU and the simulation completion signal back to CPU), GPU programs and input/output data are exchanged between CPU and the expansion card. Moreover, since on every time step of the simulation the results from the previous time step are used but not changed, the results are preferably transferred back to CPU in parallel with the computation.

Abstract: An accelerator system is implemented on an expansion card comprising a printed circuit board having (a) one or more graphics processing units (GPU), (b) two or more associated memory banks (logically or physically partitioned), (c) a specialized controller, and (d) a local bus providing signal coupling compatible with the PCI industry standards (this includes but is not limited to PCI-Express, PCI-X, USB 2.0, or functionally similar technologies). The controller handles most of the primitive operations needed to set up and control GPU computation. As a result, the computer's central processing unit (CPU) is freed from this function and is dedicated to other tasks. In this case a few controls (simulation start and stop signals from the CPU and the simulation completion signal back to CPU), GPU programs and input/output data are the information exchanged between CPU and the expansion card.

Abstract: An accelerator system is implemented on an expansion card comprising a printed circuit board having (a) one or more graphics processing units (GPUs), (b) two or more associated memory banks (logically or physically partitioned), (c) a specialized controller, and (d) a local bus providing signal coupling compatible with the PCI industry standards. The controller handles most of the primitive operations to set up and control GPU computation. Thus, the computer's central processing unit (CPU) can be dedicated to other tasks. In this case a few controls (simulation start and stop signals from the CPU and the simulation completion signal back to CPU), GPU programs and input/output data are exchanged between CPU and the expansion card. Moreover, since on every time step of the simulation the results from the previous time step are used but not changed, the results are preferably transferred back to CPU in parallel with the computation.

Abstract: An accelerator system is implemented on an expansion card comprising a printed circuit board having (a) one or more graphics processing units (GPUs), (b) two or more associated memory banks (logically or physically partitioned), (c) a specialized controller, and (d) a local bus providing signal coupling compatible with the PCI industry standards. The controller handles most of the primitive operations to set up and control GPU computation. Thus, the computer's central processing unit (CPU) can be dedicated to other tasks. In this case a few controls (simulation start and stop signals from the CPU and the simulation completion signal back to CPU), GPU programs and input/output data are exchanged between CPU and the expansion card. Moreover, since on every time step of the simulation the results from the previous time step are used but not changed, the results are preferably transferred back to CPU in parallel with the computation.