Research

Disclosed are various embodiments for computing 2-body statistics on graphics processing units (GPUs). Various types of two-body statistics (2-BS) are regarded as essential components of data analysis in many scientific and computing domains. However, the quadratic complexity of these computations hinders timely processing of data. According, various embodiments of the present disclosure involve parallel algorithms for 2-BS computation on Graphics Processing Units (GPUs). Although the typical 2-BS problems can be summarized into a straightforward parallel computing pattern, traditional wisdom from (general) parallel computing often falls short in delivering the best possible performance. Therefore, various embodiments of the present disclosure involve techniques to decompose 2-BS problems and methods for effective use of computing resources on GPUs. We also develop analytical models that guide users towards the appropriate parameters of a GPU program.

See the patent here 

A new compromise-resilient and compact cryptographic tool is provided that ensures a breach-resilient authentication and integrity of system measurements in computer systems. The described methods are forward-secure digital signatures with signature and partial public key aggregation capabilities. The methods reduce the total space overhead of signature and public key storage. The methods offer a high space efficiency for systems who has relatively low state transitions, wherein the same message is continuously signed and then followed by different messages.

See the patent here 

A computer-based system and method for generating a current pain assessment of a neonate using facial expressions along with crying sounds, body movement, and vital signs changes and for using the current pain objective assessment to predict future pain objective assessment and assign a future pain probability score by incorporation spatiotemporal data into the multimodal assessment.

See the patent here  

The present disclosure presents various systems and methods for implementing a physical unclonable function device. One such method comprises providing an integrated circuit having a plurality of set/reset flip flop logic circuits, wherein each of the set/reset flip flop logic circuits enters a metastable state for a particular input sequence. The method includes varying circuit parameters for each of the plurality of set/reset flip flop logic circuits to account for manufacturing variations in the set/reset flip flop logic circuits and enable generating a stable but random output in response to the particular input sequence. Thus, by applying the particular input sequence to the integrated circuit, a unique identifier for the integrated circuit can be derived from an output response of the plurality of set/reset flip flop logic circuits.

See the patent here  

In the past few years, in addition to the computing capacity that has grown exponentially, graphics processing units (GPUs) have undergone a dramatic evolution in hardware architecture and software environment. On the other hand, existing join algorithms are designed for earlier GPU architectures. Therefore it is not clear whether they can make the most out of latest devices in the market. Although the GPU code may scale well with the increasing amount of computing resources in newer GPU devices, maximum performance cannot be achieved without optimization towards new GPU components and features in the runtime system software.

See the patent here  

Programming a dynamical system such as a robot to do a large number of tasks under every possible set of constraints is unrealistic. Inspired by humans' remarkable imitation learning capability, researchers introduced Learning from Demonstration (LfD) or Imitation Learning (IL) which learns the policies of a dynamical system performing a task from the demonstrations performed by experts.

See the patent here  

There has been an increasing interest in the past several years to improve the current practice of assessing neonates' pain by developing automated and round-the-clock systems. The inventors propose a non-invasive and comprehensive system that simultaneously measures the traditionally observed behavioral and physiological pain responses and utilizes them to assess neonatal pain.

See the patent here  

Disclosed are various embodiments for a framework for graphics processing unit (GPU) based spatial indexing and query processing. These embodiments provide a more comprehensive spatial indexing framework with even better performance than previous frameworks. Likewise, these embodiments also support functionality beyond query processing.

See the patent here  

Handling analytical workloads efficiently is a major challenge in today's scientific domains. Recent studies show increasing interest in developing database systems for handling scientific data. Traditional database management systems (DBMSs) still fall short of algorithms and strategies to satisfy the special needs of scientific applications, which are very different from those in traditional databases in their data types and query patterns. In addition, design of efficient algorithms for data query and analysis are still the main challenge in scientific areas. In addition, support of complex mathematical functions in DBMS have become an active research area.

See the patent here 

There are close to 4500 different species of mosquitoes in the world spread across 45 or so genera. Out of these, only handfuls of species across three genus types spread the deadliest diseases. As detailed in this disclosure, digitizing anatomies of mosquito specimens across the globe (with citizen and expert involvement) will help create a massive repository of mosquito anatomy images tagged with genus and species types. This repository could then be used for training personnel, and also for automatic identification using algorithms in this disclosure (when a picture is uploaded).

See the patent here  

Design-based (unbiased) stereology is the current best practice for quantifying the number of cells in a tissue sample. The majority of funding agencies, journal editors, and regulatory bodies prefer the sound mathematical basis of stereology approaches over assumption- and model-based methods. The major obstacle to high throughput applications is that current stereology approaches require time- and labor-intensive manual data collection, which can be prohibitive on tissue samples that include multiple cell types. For example, section or slice thickness determination may be carried out by a user performing manual adjustments using the microscope's fine focusing mechanism to locate the boundaries of slice. In addition, a user may also be required to manually locate and select objects of interest while stepping through stained tissue sections in order to perform quantitative analysis of biological microstructures. Therefore, there is a continuing need to reduce the number of manual steps required, as well as increase the efficiency and accuracy of automated stereology.

See the patent here  

Tech-based solutions to combat the spread of mosquito-borne diseases is an important need of the hour. However, there is no system yet that enables common citizens to participate in mosquito identification. This disclosure fills the gap by designing a computerized process, such as one enabled in a smart-phone based system, that enables anyone to take images of a still mosquito that is alive or dead (after possibly spraying or trapping), but still retaining its physical form, and then processes the images for species identification.

See the patent here 

See the patent here  

The embodiments present a new class of content masking attacks against the Portable Document Format (PDF) standard. The attacks cause documents to appear different than the underlying content extracted from the documents by information-based services. The first attack allows academic paper writers and reviewers to collude via subverting the automatic reviewer assignment systems in current use by academic conferences. The second attack renders plagiarism detection software ineffective, targeting specific small plagiarism similarity scores to appear natural and evade detection. The final attack places masked content into the indexes for various search engines, which renders as information entirely different from the keywords used to locate it. The final attack enables spam, profane, or possibly illegal content to go unnoticed by these search engines but still be returned in search results.

See the patent here  

Infants' pain assessment is challenging since infants do not have the ability to communicate or articulate their pain experience. Assessing their pain depends primarily on the subjective judgments of nurses who monitor infants periodically and observe signs of pain [2]. Therefore, it is important to build an objective pain assessment system to measure infants' pain intensity based on physiological (e.g., vital signs, changes in skin color, and pupil dilation) and/or behavioral (e.g., pain expression, crying, and body movements) pain indicators.

See the patent here  

Disclosed is a radiological method for predicting the severity of a tumor in a subject that involves spatially superimposing two or more radiological images of the tumor sufficient to define regional habitat variations in two or more ecological dynamics in the tumor, and comparing the habitat variations to one or more controls to predict the severity of the tumor. In some cases, the method predicts the survival of the subject based on the severity of the tumor.

See the patent here  

In some embodiments, a functional object-oriented network (FOON) is provided that includes a plurality of functional units each comprising two or more input object nodes that each identify an object and its state before a manipulation motion is performed, a motion node that identifies a manipulation motion that can be performed using the objects, and one or more output object nodes that each identify an object and its state after the manipulation motion has been performed using the objects. In some embodiment, a robot can used the FOON to determine the discrete actions that are required to perform a given task.

See the patent here  

A subjective analysis of stained tissue sections is a critical step in the detection and diagnosis of most cancer in developed countries. To help differentiate cancerous from normal tissue in these decisions, stereological parameters exist to quantify mean nuclear size and 3-D patterns of clustering and anisotropy. Accordingly, there is a need in the art for a system and method that reduces the time and cost for a trained expert to manually identify possible cases of cervical and other forms of cancer based on stained tissue sections from Pap smears and cervical biopsies.

See the patent here  

In traditional public key settings, to thwart man-in-the-middle attacks, certificates are issued on user-generated public keys to ensure authenticity. However, the infrastructure needed to issue and manage certificates has shown to be very costly to establish and maintain. More importantly, the need to obtain and verify certificates (or certificate chains) can incur substantial computational overhead. For instance, certificate authorities must maintain active or expired certificates, issue new certificates, and verify certificates. While computational overhead in issuing, maintaining, and verifying certificates might be tolerable for certain hardware, it can be non-ideal for internet-of-things (IoT) devices where heterogeneous low-end devices need to communicate securely with minimal use of computational resources.

See the patent here  

System and method for processing, non-concurrently collected, electroencephalogram (EEG) data and resting station functional magnetic resonance imaging (rsfMRI) data, non-invasively, to create a patient-specific three-dimensional (3D) mapping of the patient's functional brain network. The mapping can be used to more precisely identify candidates of resective neurosurgery and to help create a targeted surgical plan for those patients. The methodology automatically maps the patient's unique brain network using non-concurrent EEG and resting state functional MRI (rsfMRI). Generally, the current invention merges non-concurrent EEG data and rsfMRI data to map the patient's epilepsy/seizure network.

See the patent here  

A system and methodology for the automated localization, extraction, and analysis of MRI-based features with clinical information to improve the diagnosis of pelvic organ prolapse (POP). The system can automatically identify reference points for pelvic floor measurements on MRI rapidly and consistent. It provides a prediction model that analyzes the correlation between current and new MRI-based features with clinical information to differentiate patients with and without POP. This system will enable the high throughput analysis of MR images for their correlation with clinical information to better detect POP. The presented system can also be applied to the automated localization and extraction of MRI features for the diagnosis of other diseases where clinical examination is not adequate.

See the patent here 

A computer data security system, useful in protecting audit logs, includes symmetric key based techniques, requires only a small-constant number of cryptographic hash operations at the signer side sending a prospective audit log or other computer record data to a primary repository to achieve forward-secure and append-only authentication. The verification is performed by independent parties sharing parts of the symmetric key, wherein the presence of single honest party among all verifier parties ensures a conditional non-repudiation. It also ensures that an active adversary cannot generate authentication tags on behalf of the signer, unless it compromises all verification parties.

See the patent here 

In the context of real-world mobile robots, state estimation is the problem of estimating the state of a robot. Typically, states such as the exact location of a robot and the exact locations of obstacles in a robot's environment are not directly observable. However, such states can be inferred from sensor measurements. A robot can rely on sensor measurements to infer its state and the state of its environment. Unfortunately, sensor measurements can be noisy and the amount of noise can vary with state. Various methods of estimating and iteratively adapting measurement noise over time can be performed. However, those methods do not assume that measurement noise is stochastic and they do not estimate sensor measurement bias and noise based on state. It would be preferable to estimate noise and bias based on state of the robot.

See the patent here 

A digital signature protocol achieves the least energy consumption, the fastest signature generation, and the smallest signature among its ECC-based counterparts. The method of authenticating is also immune to side channel attacks aiming elliptical curve operations/exponentiations as well as to weak pseudo random number generators at the signer's side. A security system for authenticating the updating of computer records includes a network of member computers in data communication with each other and calculating an assigned portion of a commitment code used for the authenticating the secure data at the verifier device by completing, among the server computers, a plurality of discrete verification tasks for a single authentication code (R−) associated with an aggregated commitment code (R), aggregating at the verifier device, the aggregated commitment code (R), and verifying the secure data at the verifier device.

See the patent here 

A method and system for generating structured data from semi-structured data are provided. The method includes reading a plurality of records from a data file including semi-structured data. Further, the method includes obtaining aligned delimiters in a list for every record that has been read. The method also includes selecting a most occurring delimiter from the list. The method then includes constructing a regular expression using the selected delimiter to split the records into different fields. The method also includes reconstructing the records for the regular expression to fit and split into fields. In addition, the method includes displaying the records split into the fields.

See the patent here 

A system and method for identifying a subject based upon ear recognition using a convolutional neural network (CNN) and handcrafted features, wherein an ear in an image is cropped using ground truth annotations and landmark detection is performed to obtain the information required to normalize pose and scale variations. The normalized images are then described by different feature extractors and matched through distance metrics. Finally, scores are fused and a subject identification decision is made.

See the patent here 

Lung cancer is responsible for the greatest number of cancer related deaths in the United States. The advent of low-dose helical computed tomography has made it possible to provide relatively low risk screening for high risk patients. Though still somewhat controversial, results of the National Lung Screening Trial (NLST) have shown a 20% decrease in mortality with the use of low-dose CT compared to X-ray findings. The sensitivity of the procedure is its bane; many of the detected nodules are not cancerous. Accurate classification (or prediction) of pulmonary nodules to be cancerous is key to determining further diagnosis and treatment options.

See the patent here 

Generally, gait is the pattern that the body frame and/or limbs of animals exhibit during movement. Most animals exhibit different gaits based on certain factors, such as speed and terrain, among others. Human gait refers to the pattern or patterns exhibited by the movement of the human body frame and/or human limbs. For example, human gait may be described in terms of bipedal forward propulsion of the human body, including alternate but repeating movements of different body segments. Different gait patterns are characterized by differences in limb movement patterns, such as differences in velocity, forces, and potential energy, as well as differences due to changes in surface contact. Thus, different human gaits are representative of the different ways that humans can move.

See the patent here 

A system for generating symmetric cryptographic keys for communications between hosts. Hosts use associated devices to generate secret keys. Each key is generated based on a static seed and a dynamic seed. The dynamic seed is created from sensor data or auxiliary data. The secret key allows host machines to encrypt, or decrypt, plaintext messages sent to, or received from, other host machines.

See the patent here 

A magnetic system for solving one or more quadratic optimization problems by associating each of a plurality of variables of a quadratic optimization problem with a nanomagnet subset of a nanomagnet array, driving the nanomagnets of the nanomagnet subset to an excited state, allowing the nanomagnets of the nanomagnet subset to enter a relaxed state after being driven to an excited state, wherein the nanomagnets magnetically couple with one another in the relaxed state to minimize the total magnetic coupling energy of the nanomagnet array, and sensing a magnetic coupling of the nanomagnets of the nanomagnet subset to solve the quadratic optimization problem.

See the patent here 

Various devices, methods and systems are provided for aging-sensitive chip authentication. In one example, among others, a chip includes a reference Schmitt trigger ring oscillator (STRO) configured to enter a sleep mode during operation of the chip; a stressed STRO; a VDD charge pump configured to boost a positive voltage supplied to the stressed STRO during operation of the chip; and a GND charge pump configured to under-drive a ground voltage supplied to the stressed STRO during operation of the chip. In another example, a method includes detecting activation of a chip including a reference STRO and a stressed STRO and, in response to the activation of the chip, initiating sleep mode operation of the reference STRO. In response to the activation of the chip, a VDD voltage supplied to the stressed STRO can be boosted and/or a GND voltage supplied to the stressed STRO can be under-driven.

See the patent here 

Aspects of wide operating range level shifter designs are described. One embodiment includes a level shifter configured to receive an input signal in a first voltage domain and generate an output signal in a second voltage domain, a pulse generator configured to generate a pulse in response to sensing a rise transition on the input signal, and a droop circuit configured to decouple at least a portion of the level shifter from the second voltage domain in response to the pulse. According to one aspect of the embodiments, the pulse can be provided to the droop circuit to decouple at least a portion of the level shifter from the second voltage domain and reduce contention between transistors in the level shifter. Using the concepts described herein, the worst case rise time delay for level shifters can be significantly reduced.

See the patent here 

This invention relates to a system and method for applying an ensemble of segmentations to microscopy images of a tissue sample to determine if the tissue sample is representative of cancerous tissue. The ensemble of segmentations is applied to a plurality of greyscale or color microscopy images to generate a final image level segmentation and a final blob level segmentation. The final image level segmentation and final blob level segmentation are used to calculate a mean nuclear volume to determine if the tissue sample is representative of cancerous tissue.

See the patent here 

A system and method for providing a physically unclonable function (PFU) is described. In operation, the method includes applying a domain wall shift pulse challenge to a plurality of nanowires of a domain wall memory (DWM) array, wherein the nanowires of the domain wall memory (DWM) array have process induced variations, resulting in pinning potentials which affect the velocity of the domain walls along the length of the nanowires. Following the application of the domain wall shift pulse, the response to the challenge is determined by measuring the response of the plurality of nanowires of the domain wall memory to the applied domain wall shift pulse challenge to provide a physically unclonable function (PUF) for the integrated circuit.

See the patent here