SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
We present a user interface for software reuse repository that relies both on the informal semantics of faceted classification and the formal semantics of type signatures for abstract data types. The result is an interface providing both structural and qualitative feedback to a software reuser.
Cardiac magnetic resonance imaging provides high spatial resolution, enabling improved extraction of important functional and morphological features for cardiovascular disease staging. Segmentation of ventricular cavities and myocardium in cardiac cine sequencing provides a basis to quantify cardiac measures such as ejection fraction. A method is presented that curtails the expense and observer bias of manual cardiac evaluation by combining semantic segmentation and disease classification into a fully automatic processing pipeline. The initial processing element consists of a robust dilated convolutional neural network architecture for voxel-wise segmentation of the myocardium and ventricular cavities. The resulting comprehensive volumetric feature matrix captures diagnostic clinical procedure data and is utilized by the final processing element to model a cardiac pathology classifier. Our approach evaluated anonymized cardiac images from a training data set of 100 patients (4 pathology groups, 1 healthy group, 20 patients per group) examined at the University Hospital of Dijon. The top average Dice index scores achieved were 0.940, 0.886, 0.849 for structure segmentation of the left ventricle (LV), myocardium and right ventricle (RV) respectively. A 5-ary pathology classification accuracy of 90% was recorded on an independent test set using the trained model. Performance results demonstrate potential for advanced machine learning methods to deliver accurate, efficient and reproducible cardiac pathological assessment.
The paper deals with some design issues involved in building a system that could visually represent the semantic structures of training materials and their underlying mental models. In particular, hypermedia-based semantic networks that instantiate classification problem solving strategies are thought to be a useful formalism for such representations; the complexity of these web structures can be best managed through visual depictions. It is also noted that a useful approach to implement in these hypermedia models would be some metrics of conceptual distance.
We present a pipeline method that curtails the expense and observer bias of manual cardiac evaluation by combining semantic segmentation and disease classification as a fully automatic processing pipeline. The initial element consists of a 2D U-Net convolutional neural network architecture for voxel-wise segmentation of the myocardium and ventricular cavities. The results of the segmentation were used to compute a comprehensive volumetric feature matrix that captured diagnostic clinical procedure data and that was used to model a cardiac pathology classifier.Our approach evaluated anonymized parasternal MRI cardiac images from a database of 100 patients (4 pathology groups, 1 healthy group, 20 patients per group) examined at the University Hospital of Dijon. We achieved top average Dice index scores of 0.939, 0.849, 0.886 for structure segmentation of the left ventricle (LV), right ventricle (RV) and myocardium respectively. A 5-ary pathology classification accuracy of 90% was recorded on an independent test set using our trained model.
Deep learning has quickly become a necessity for selfdriving vehicles on Earth. In contrast, the self-driving vehicles on Mars, including NASA’s latest rover, Perseverance, which is planned to land on Mars in February 2021, are still driven by classical machine vision systems. Deep learning capabilities, such as semantic segmentation and object recognition, would substantially benefit the safety and productivity of ongoing and future missions to the red planet. To this end, we created the first large-scale dataset, AI4Mars, for training and validating terrain classification models for Mars, consisting of ~326K semantic segmentation full image labels on 35K images from Curiosity, Opportunity, and Spirit rovers, collected through crowdsourcing. Each image was labeled by ~10 people to ensure greater quality and agreement of the crowdsourced labels. It also includes ~1.5K validation labels annotated by the rover planners and scientists from NASA’s MSL (Mars Science Laboratory) mission, which operates the Curiosity rover, and MER (Mars Exploration Rovers) mission, which operated the Spirit and Opportunity rovers. We trained a DeepLabv3 model on the AI4Mars training dataset and achieved over 96% overall classification accuracy on the test set. The dataset is made publicly available.1
Computer supported indexing systems may be categorized in several ways. One classification scheme refers to them as statistical, syntactic, semantic or knowledge-based. While a system may emphasize one of these aspects, most systems actually combine two or more of these mechanisms to maximize system efficiency. Statistical systems can be based on counts of words or word stems, statistical association, and correlation techniques that assign weights to word locations or provide lexical disambiguation, calculations regarding the likelihood of word co-occurrences, clustering of word stems and transformations, or any other computational method used to identify pertinent terms. If words are counted, the ones of median frequency become candidate index terms. Syntactical systems stress grammar and identify parts of speech. Concepts found in designated grammatical combinations, such as noun phrases, generate the suggested terms. Semantic systems are concerned with the context sensitivity of words in text. The primary goal of this type of indexing is to identify without regard to syntax the subject matter and the context-bearing words in the text being indexed. Knowledge-based systems provide a conceptual network that goes past thesaurus or equivalent relationships to knowing (e.g., in the National Library of Medicine (NLM) system) that because the tibia is part of the leg, a document relating to injuries to the tibia should he indexed to LEG INJURIES, not the broader MeSH term INJURIES, or knowing that the term FEMALE should automatically be added when the term PREGNANCY is assigned, and also that the indexer should be prompted to add either HUMAN or ANIMAL. Another way of categorizing indexing systems is to identify them as producing either assigned- or derived-term indexes.
Data base management systems (DBMSs) are in widespread use because of the ease and flexibility with which users access large volumes of data. Ensuring data accuracy through integrity constraints is a central aspect of DBMS use. However, many DBMSs still lack adequate integrity support. In additon, a comprehensive theoretical basis for such support the role of a constraint classification system - has yet to be developed. This paper presents a formalism that classifies semantic integrity constraints based on the structure of the relational model. Integrity constraints are characterized by the portion of the data base structure they access, whether one or more relations, attributes, or tuples. Thus, the model is completely general, allowing the arbitrary specification of any constraint. Examples of each type of constraint are illustrated using a small engineering data base, and various implementation issues are discussed.
Large-language models (LLMs) have been recently used for spoken language understanding (SLU) to infer meaning and semantics from speech in tasks such as speaker intent and sentiment classification. Due to being trained on large amounts of data, and their ability to understand context and relationships between words, LLMs are competent, enabling them to generalize across tasks without requiring many task-specific training samples. This research examines the feasibility of few-shot learning in LLMs for detecting subtle, brief, and possibly unconscious interactions between team members, called ``micro-behaviors," and provides insights into the appropriate design of LLMs for this task. Our data came from 5 teams participating in a 45-day mission at the US National Aeronautics and Space Administration’s (NASA) Human Exploration Research Analog (HERA). More specifically we used data collected from team interaction battery (TIB) tasks teams performed five times in-mission which comprise an average 1.5 hours of conversation data per day. Micro-behaviors were coded according to an adapted version of Smith & Griffins (2022) theoretical framework in terms of Violation (i.e., presence of valenced behavior, uplifting/positive or discouraging/negative), Intensity (i.e., force of behavior in terms of how uplifting or discouraging is the behavior), and Intent (i.e., motive of the behavior in terms of whether it was deliberate or unintentional). We explore the ability of LLMs to detect the presence and intensity of micro-behaviors. We examine employing and fine-tuning readily available LLMs (i.e., RoBERTa, DistilBERT), as well as prompting state-of-the-art sequence classification models (i.e., Llama-2, Llama-3). In a total of 13,058 conversational turns (17.8% uplifting, 3.3% discouraging, 75.76% neutral, 3.14% nulls), we compute the macro F1-score of the 3-way micro-behavior classification task (i.e., classifying among uplifting, discouraging, and neutral; 33% chance). Results indicate that the RoBERTa model achieves a F1-score of 36.2% (uplift: 43.3% precision (P), 15.1% recall (R); discourage: 20% P, 0.5% R). These results significantly improve when we augment the data via paraphrasing in the RoBERTa model, reaching a 41.2% macro F1-score (uplift: 37.7% P, 86.3% R; discourage: 3.5% P, 1.8% R). Finally, the Llama-2 model with 3-shot prompting yields 38% macro F1-score (uplift: 28.7% P, 20% R; discourage: 7.2% P, 18% R), which is slightly better compared to the RoBERTa model without data augmentation, highlighting the effectiveness of sequence classification models in detecting minority classes with a small sample size. Findings indicate that LLMs hold potential to detect subtle behaviors in conversations, which could be valuable in assessing team behavior in space exploration missions. Future studies will evaluate the performance of different LLM prompting strategies or fine-tuning methods.
The looming threat orbital debris poses to assets in orbit demands solutions. As the orbital population grows, so does this hazard, but so does the sea of data. The problem is also an opportunity for interdisciplinary innovation and cooperation. This paper focuses on the data and information management aspect of developing solutions for a sustainable and safe orbital space environment. The corresponding author’s in-progress work to develop an orbital debris domain ontology is summarized in order to discuss knowledge modeling for this domain. Methodological approaches of this effort can also contribute to standards efforts and address terminological and policy questions. Leveraging the growing volumes of orbital debris and space situational awareness (SSA) data will create a more complete picture of the orbital space environment. Part of the solution will be: consistent and correct data interpretation, sharing orbital debris and SSA data in one form or another, terminology development & harmonization, and knowledge or domain modeling. To facilitate this, [Rovetto, 2015/16] discussed ontology development for the orbital debris domain. This paper lists concepts from that paper, and subsequently developed concepts [2-9]. Ontology engineering is an interdisciplinary field related to knowledge representation and reasoning in artificial intelligence, semantic technologies and the so-called semantic web. An ontology is effectively a computable and semantically rich terminology that presents a knowledge or domain model for a topic area. Expressions of knowledge or assertions are stored using formally defined term. This knowledge base is reasoned over to yield answers to queries, among other things. Ontologies have been developed in knowledge-based projects across various disciplines, and used for such things as search engines, chatbots, enterprise knowledge graphs, etc. Ontologies support: interoperability, automated reasoning, data sharing and integration, data search and retrieval, and communicating the meaning of data. The Orbital Debris Ontology (ODO), and related ontologies [Rovetto & Kelso 2016] [Rovetto 2016, 2017], were proposed to help achieve this. ODO, for instance, is intended as a domain ontology that can be used across federated databases, offering an explicitly specified set of concepts describing the orbital debris domain. Its meaning-rich taxonomy will provide a sharable semantics for orbital debris data to, in part, consistently communicate the meaning of data to both humans and machines, and tag data elements in space object catalogs to help afford inference tasks, decision support, knowledge discovery, and information integration. ODO and the SSA ontology (SSAO) is part of the overall Orbital Space Domain Ontology concept, which is conceived as a broader domain reference ontology. It aims to provide a knowledge representation structure of the orbital space environment, a common semantic model, and develop a sharable terminology. Collectively this will provide common meaning for datasets, a high-level taxonomy or classification for orbital space objects, and thus means to characterize space objects. Ongoing efforts have included using visualizations, R, JSON-LD, and contemporary semantic technologies. Potential applications and interdisciplinary partnerships include web-based platforms, web apps, visualizations, and academia projects. Community input and participation may yield a more widely understood domain model as well as facilitate terminological standards. For example, the proposed conceptual, terminological and ontological analysis may contribute to such efforts as the Space Debris Mitigation Requirements in the International Standards Organization by developing more precise, consistent and coherent terms and definitions. Projects that seek to develop in-house ontologies can use ODO and related ontologies as domain reference ontologies. This paper was developed independent of author affiliations. Readers are encouraged to contact corresponding author(1) with general interest and potential opportunities to support or realize the described project.
This paper addresses an unconventional application of a production system to a problem involving belief specialization. The production system reduces a large quantity of low-level descriptions into just a few higher-level descriptions that encompass the problem space in a more tractable fashion. This classification process utilizes a set of descriptions generated by combining the component hierarchy of a physical system with the semantics of the terminology employed in its operation. The paper describes an application of this process in a program, constructed in C and CLIPS, that classifies signatures of electromechanical system configurations. The program compares two independent classifications, describing the actual and expected system configurations, in order to generate a set of contradictions between the two.
Image segmentation involves labelling pixels according to their membership in image regions. This requires the understanding of what a region is. Using supervised pixel classification, the paper investigates how groups of pixels labelled manually according to perceived image semantics map onto the feature space created by an Artificial Visual System. Multiscale structure of regions are investigated and it is shown that pixels form clusters based on their geometric roles in the image intensity function, not by image semantics. A tentative abstract definition of a 'region' is proposed based on this behavior.
We present NeMO-Net, the Srst open-source deep convolutional neural network (CNN) and interactive learning and training software aimed at assessing the present and past dynamics of coral reef ecosystems through habitat mapping into 10 biological and physical classes. Shallow marine systems, particularly coral reefs, are under significant pressures due to climate change, ocean acidification, and other anthropogenic pressures, leading to rapid, often devastating changes, in these fragile and diverse ecosystems. Historically, remote sensing of shallow marine habitats has been limited to meter-scale imagery due to the optical effects of ocean wave distortion, refraction, and optical attenuation. NeMO-Net combines 3D cm-scale distortion-free imagery captured using NASA FluidCam and Fluid lensing remote sensing technology with low resolution airborne and spaceborne datasets of varying spatial resolutions, spectral spaces, calibrations, and temporal cadence in a supercomputer-based machine learning framework. NeMO-Net augments and improves the benthic habitat classification accuracy of low-resolution datasets across large geographic ad temporal scales using high-resolution training data from FluidCam.NeMO-Net uses fully convolutional networks based upon ResNet and ReSneNet to perform semantic segmentation of remote sensing imagery of shallow marine systems captured by drones, aircraft, and satellites, including WorldView and Sentinel. Deep Laplacian Pyramid Super-Resolution Networks (LapSRN) alongside Domain Adversarial Neural Networks (DANNs) are used to reconstruct high resolution information from low resolution imagery, and to recognize domain-invariant features across datasets from multiple platforms to achieve high classification accuracies, overcoming inter-sensor spatial, spectral and temporal variations.Finally, we share our online active learning and citizen science platform, which allows users to provide interactive training data for NeMO-Net in 2D and 3D, integrated within a deep learning framework. We present results from the PaciSc Islands including Fiji, Guam and Peros Banhos 1 1 2 1 3 1 where 24-class classification accuracy exceeds 91%.
Expert System shells are lacking in many areas of software engineering. Large rule based systems are not semantically comprehensible, difficult to debug, and impossible to modify or validate. Partitioning a set of rules found in CLIPS (C Language Integrated Production System) into groups of rules which reflect the underlying semantic subdomains of the problem, will address adequately the concerns stated above. Techniques are introduced to structure a CLIPS rule base into groups of rules that inherently have common semantic information. The concepts involved are imported from the field of A.I., Pattern Recognition, and Statistical Inference. Techniques focus on the areas of feature selection, classification, and a criteria of how 'good' the classification technique is, based on Bayesian Decision Theory. A variety of distance metrics are discussed for measuring the 'closeness' of CLIPS rules and various Nearest Neighbor classification algorithms are described based on the above metric.
This paper describes an investigation into the structure of representations of sets of actions, utilizing semigroup theory. The goals of this project are twofold: to shed light on the relationship between tasks and representations, leading to a classification of tasks according to the representations they admit; and to develop techniques for automatically transforming representations so as to improve problem-solving performance. A method is demonstrated for automatically generating serial algorithms for representations whose actions form a finite group. This method is then extended to representations whose actions form a finite inverse semigroup.
NASA NeMO-Net, the neural multi-modal observation and training network for global coral reef assessment, is an open-source deep convolutional neural network and interactive active learning training software aiming to accurately assess the present and past dynamics of coral reef ecosystems through determination of percent living cover and morphology as well as mapping of spatial distribution. We present an interactive citizen science video game, released this April, for desktop and iOS devices where users interactively label morphology classifications over mm-scale 3D coral reef imagery captured using diver photomosaic imagery, the UAV enabled NASA FluidCam instrument, and satellite datasets. To date, the application has had over 40,000 downloads and over 60,000 unique coral reef classifications, each filtered through a user-based rating and expert evaluation system. We also present results from NeMO-Net’s convolutional neural network (CNN) models used to semantically segment 2D satellite imagery as well as projections of 3D coral reconstructions using user input data as training datasets. Fusing datasets using machine learning from multiple remote sensing platforms presents novel methodologies for assessing the health of coral ecosystems, which are critically endangered by a changing climate. In partnering with Mission Blue, the National Oceanic and Atmospheric Administration (NOAA), and the Living Oceans Foundation (LOF), NeMO-Net leverages an international consortium of subject matter experts to provide both proper training for citizen scientists and the generation of a labeled datasets to ingest into machine learning algorithms for global coral reef identification.
NASA NeMO-Net, the neural multi-modal observation and training network for global coral reef assessment, is an open-source deep convolutional neural network and interactive active learning training software aiming to accurately assess the present and past dynamics of coral reef ecosystems through determination of percent living cover and morphology as well as mapping of spatial distribution. We present an interactive citizen science video game, released this April, for desktop and iOS devices where users interactively label morphology classifications over mm-scale 3D coral reef imagery captured using diver photomosaic imagery, the UAV enabled NASA FluidCam instrument, and satellite datasets. To date, the application has had over 40,000 downloads and over60,000 unique coral reef classifications, each filtered through a user-based rating and expert evaluation system. We also present results from NeMO-Net’s convolutional neural network (CNN) models used to semantically segment 2D satellite imagery as well as projections of 3D coral reconstructions using user input data as training datasets. Fusing datasets using machine learning from multiple remote sensing platforms presents novel methodologies for assessing the health of coral ecosystems, which are critically endangered by a changing climate. In partnering with Mission Blue, the National Oceanic and Atmospheric Administration (NOAA), and the Living Oceans Foundation (LOF), NeMO-Net leverages an international consortium of subject matter experts to provide both proper training for citizen scientists and the generation of a labeled datasets to ingest into machine learning algorithms for global coral reef identification.
Clarissa, an experimental voice enabled procedure browser that has recently been deployed on the International Space Station, is as far as we know the first spoken dialog system in space. We describe the objectives of the Clarissa project and the system's architecture. In particular, we focus on three key problems: grammar-based speech recognition using the Regulus toolkit; methods for open mic speech recognition; and robust side-effect free dialogue management for handling undos, corrections and confirmations. We first describe the grammar-based recogniser we have build using Regulus, and report experiments where we compare it against a class N-gram recogniser trained off the same 3297 utterance dataset. We obtained a 15% relative improvement in WER and a 37% improvement in semantic error rate. The grammar-based recogniser moreover outperforms the class N-gram version for utterances of all lengths from 1 to 9 words inclusive. The central problem in building an open-mic speech recognition system is being able to distinguish between commands directed at the system, and other material (cross-talk), which should be rejected. Most spoken dialogue systems make the accept/reject decision by applying a threshold to the recognition confidence score. NASA shows how a simple and general method, based on standard approaches to document classification using Support Vector Machines, can give substantially better performance, and report experiments showing a relative reduction in the task-level error rate by about 25% compared to the baseline confidence threshold method. Finally, we describe a general side-effect free dialogue management architecture that we have implemented in Clarissa, which extends the "update semantics'' framework by including task as well as dialogue information in the information state. We show that this enables elegant treatments of several dialogue management problems, including corrections, confirmations, querying of the environment, and regression testing.
As NASA's Human Research Program (HRP) prepares for long-duration Mars missions, understanding astronaut tasks is crucial. This study, conducted at NASA Glenn Research Center (GRC), employed Natural Language Processing (NLP) and machine learning techniques to analyze and classify Mars tasks. A list of 1,058 Mars tasks was provided by HRP experts including binary labeling of 18 Human System Task Categories (HSTCs). We developed an NLP model using Google's BERT language model to capture the semantic and syntactic nuances of these tasks. Supervised training was initially applied to a subset of the NLP-analyzed tasks to assess the model's effectiveness in classifying the remaining tasks. Incorporating HSTC descriptions significantly enhanced the classification accuracy for 9 out of the 18 HSTCs and reduced training time. To address the issue of severe class imbalance in the HSTC data, we introduced innovative weighting and sampling techniques for data augmentation. We then fine-tune BERT to implement a pairwise relatedness scoring method, allowing us to cluster tasks based on their relatedness and similarity, getting a step closer to labeling the tasks without supervision. In this presentation we guide you through data preprocessing, deciphering key syntax components using BERT, and performing supervised classification of the Mars tasks. This work showcases the potential use of advanced NLP techniques to analyze Mars missions to be incorporated into various crew health and performance analyses.