68 results on '"Uttam Majumder"'
Search Results
2. Influence of sociodemographic factors, diagnostic variations, and phenomenology toward the treatment response in adolescent catatonia in a tertiary care centre in Eastern India
- Author
-
Uttam Majumder, Parthasarathy Biswas, Jagadish Biswas, and Avik Kumar Layek
- Subjects
adolescence ,catatonia ,predictors ,phenomenology ,treatment response ,Medicine - Abstract
Background: Catatonia remains an area of lesser research especially in the adolescent age group. It has subtle differences in the presentation and background diagnoses as compared to adult catatonia. There is paucity of literatures regarding the role of different sociodemographic and clinical factors attributing to different treatment response. Aims and Objectives: The aim of the study was to assess the association of socio-demographic features, background psychiatric diagnoses, and different catatonic symptoms with treatment outcome of adolescent catatonia. Materials and Methods: The study considered 10–19 years old patients admitted in the in-patient department as per diagnostic and statistical manual diagnosis. They were assessed by Pediatric Catatonia Rating Scale (PCRS) and treated with lorazepam initially with varying dosage and duration. Modified electroconvulsive therapy was administered in resistance. Factors of these two groups were statistically analyzed to assess predictability towards outcome. Results: There were 47 participants with mean age of 16.66±1.21 years of whom 29.8% showed positive family history of different psychiatric illnesses. Most of them came with schizophrenia and related disorders (53.2%), though mood disorders, conversion and organic brain diseases were also there. Among them 30 (63.8%) patients responded to lorazepam treatment. Positive family history, urban background, and catatonic severity in terms of higher PCRS score showed predictability to lorazepam non-response. Clinical features such as stupor, staring, negativism, withdrawal, mutism, urinary incontinence and refusal to eat or drink were associated with non-response, whereas waxy flexibility, stereotypy, verbigeration, and mannerism were seen in the response group. Conclusion: There is need to identify warning signs such as family history, greater symptom load and certain clinical features that can lead to resistance to the treatment with benzodiazepines based on this study. It can necessitate further large-scale study to alleviate disease burden to this young and productive population.
- Published
- 2023
- Full Text
- View/download PDF
3. Cognitive deterioration in association with teeth loss in elderly population: A study from Eastern India
- Author
-
Uttam Majumder, Iti Baidya, Avik Kumar Layek, Sampa Ray Bhattacharya, and Pradip Kumar Ray
- Subjects
chewing ability ,cognitive impairment ,elderly ,mini-mental state examination ,tooth loss ,Medicine - Abstract
Background: With the global increase in cognitive impairment and dementia, the need to investigate into the possible ways that can be used to prevent or delay such occurrence has been growing. Maintenance of dental care and oral hygiene has been promoted as one such aspect. Aims and Objectives: To study the association between cognitive impairment in patients without registered diagnosis of dementia with tooth loss and chewing ability. Materials and Methods: It was a cross-sectional hospital-based study where consenting elderly persons attending Dental outpatient department over 1½ years were included. Data obtained on socio-demographic details, number of tooth loss, number of remaining teeth, subjective chewing ability, cognitive assessment as per mini-mental state examination (MMSE), and Clock Drawing Test were analyzed statistically to check correlational association. Results: In this study 96 eligible elderly persons of mean age of 68.30±6.28 years showed increased cognitive impairment among females (P=0.003), increased age (P=0.009), rural background (P=0.033) and low income groups (P=0.001). Positive correlation was found between chewing capacity (P=0.348), number of remaining teeth (r=0.418) with MMSE scores. Conclusion: Our study population showed positive correlation between impaired cognition and the number of extracted tooth and chewing capacity. With further study on wider and representative population, we hope to project the role of maintaining good oral hygiene and dental care as a possible preventive strategy among many others to combat the increased burden of cognitive impairment.
- Published
- 2022
- Full Text
- View/download PDF
4. Study on the undergraduate interns of a medical college in eastern India: the knowledge, attitude and practice regarding COVID-19 pandemic preparation
- Author
-
Uttam Majumder, Iti Baidya, and Avik Kumar Layek
- Subjects
knowledge ,attitude ,practice ,medical interns ,covid-19 ,Medicine - Abstract
Background: With the rapid propagation of the SARS-COV-2 or COVID-19 pandemic, the healthcare facility around the world has been stretched a large extent. To keep the green supply chain to the healthcare resources ready for the battle, sufficient knowledge, proper attitude and adequate practices are of paramount necessity. Aims and Objectives: The study was done to the knowledge, practice and attitudes towards COVID-19 pandemic preparedness among the undergraduate medical interns of a tertiary care teaching center in Eastern India. Materials and Methods: Semi-structured proforma for socio-demographic details and KAP Questionnaire for COVID-19 preparedness were circulated among the Interns. Interaction among the subsets of the KAP scale as well as their association with different socio-demographic variables were studied. Results: Out of the 138 subjects participated in the study the KAP parameters were not significantly different based on the socio-demographic factors barring presence of better practices among the female interns and less score on knowledge and attitude in presence of Psychiatric illness. It was seen that better practice was significantly correlated with knowledge mean score. Conclusion: The study implied that training on the updated knowledge along with exposure to simulated environment with scheduled supervision to reflect the behavior of the interns was of great importance so that in extreme situation, the less experienced resources could also come handy into utilization if needed.
- Published
- 2021
- Full Text
- View/download PDF
5. Toward Near-Real-Time Training With Semi-Random Deep Neural Networks and Tensor-Train Decomposition
- Author
-
Humza Syed, Ryan Bryla, Uttam Majumder, and Dhireesha Kudithipudi
- Subjects
Convolutional neural networks (CNNs) ,extreme learning machine (ELM) ,image classification ,neural networks with random weights ,random vector functional link (RVFL) ,synthetic aperture radar (SAR) ,Ocean engineering ,TC1501-1800 ,Geophysics. Cosmic physics ,QC801-809 - Abstract
In recent years, deep neural networks have shown to achieve state-of-the-art performance on several classification and prediction tasks. However, these networks demand undesirable lengthy training times coupled with high computational resources (memory, I/O, processing time). In this work, we explore semi-random deep neural networks to achieve near real-time training and less computational resource usage. Although many works enhance the underlying hardware for real-time training, this work focuses on algorithmic optimization. It is shown that random projection networks with additional skipped connectivity and randomly weighted layers can boost the overall network performance while enabling for real-time training. Additionally, a tensor-train decomposition technique is leveraged to further reduce the model complexity of these networks. Our investigation accomplishes the following: 1) Tensor-train decomposition decreases the complexity of random projection networks, 2) compression of the fully connected hidden layer leads to a minimum $\sim40\times$ decrease in memory size, and 3) training under random projection networks can be achieved in near-real time.
- Published
- 2021
- Full Text
- View/download PDF
6. Correlational Analysis of Socio-demographic and Clinical Profile in Determining the Treatment Response in Patients with Catatonia in the Psychiatric Inpatient Department of a Rural Tertiary Care Centre in Eastern India
- Author
-
Avik Kumar Layek, Uttam Majumder, and Parthasarathy Biswas
- Subjects
electroconvulsive therapy ,immobility ,lorazepam ,mutism ,waxy flexibility ,Medicine - Abstract
Introduction: Catatonia, a poorly understood syndrome challen-ging the clinician’s diagnostic and management skills, has scarce literatures regarding the clinical correlates and determining factors towards treatment response. Aim: To find the correlates of socio-demographic, clinical profile, catatonic features and identifying determining factors of treatment response to lorazepam and Modified Electroconvulsive Therapy (MECT) in catatonia. Materials and Methods: This cross-sectional study was conducted at the Department of Psychiatry, North Bengal Medical College, Siliguri, West Bengal, India from January 2020 to February 2021. The catatonia cases satisfying the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria were studied. A total of 66 patients were evaluated using the 23-item Bush Francis Catatonia Rating Scale (BFCRS) for severity and later grouped into lorazepam responder (Group I) and non responders (Group II) who received MECT. Background diagnoses using DSM-5 was made after symptom resolution. Statistical analyses like Chi-square and student’s t-test to compare frequencies and means respectively, Pearson’s and Spearman’s correlation test for bivariate correlation and linear and logistic regression to predict factors for treatment outcome were employed using Statistical Package for the Social Sciences (SPSS) 16.0 with a p-value
- Published
- 2022
- Full Text
- View/download PDF
7. Prevalence of Psychiatric Morbidity Among Undergraduate Students of a Dental College in West Bengal
- Author
-
Manabendra Makhal, Pradip Kumar Ray, Sampa Ray (Bhattacharya), Subhankar Ghosh, Uttam Majumder, Shantanu De, Gautam Kumar Bandyopadhyay, and Nirmal Kumar Bera
- Subjects
academic achievement ,ghq-28 ,mental health ,Medicine - Abstract
Background: Stresses of medical course affects the academic performance as well as physical and psychological health of the students negatively. In the present day scenario every one of us has vulnerability to have psychiatric morbidity. Like other young adults, undergraduate dental students are similarly vulnerable to turmoil due to academic and social stresses, which often hamper the educational achievement. Early detection and treatment of psychological morbidities shorten the suffering leading to less social impairment in long term. Aim: The aim of this study was to assess the psychiatric morbidity of dental students and the factors affecting their mental health. Settings and Design: This cross sectional, descriptive and correlation study was conducted in North Bengal Dental College, a rural dental college of West Bengal, India. Materials and Methods: The study sample consisted of a total of 89 dental students. The student enrollment was done by “simple random sampling method”. The semi-structured proforma and the General Health Questionnaire (GHQ)-28 were used to collect data and to assess the psychiatric morbidity. Statistical Analysis: Pearson’s correlation followed by multivariate linear regression analysis was done to assess the effect of academic achievement, positive and negative event on the GHQ total score. Results: The overall mean GHQ total score in the study population was 5.33, with a SD of 4.85 which was above the cutoff (>4) score. Karl Pearson correlation co-efficient r-values for GHQ total score with ‘academic achievement’ and ‘negative events’ score were 0.718 and 0.628 respectively and p-values were 0.001 and 0.001. Conclusion: The poor academic performance and negative events had strong impact on psychological morbidity of dental students. The findings of our study can’t be generalized for all undergraduate dental student due to small sample size and non inclusion of many other variables.
- Published
- 2015
- Full Text
- View/download PDF
8. Joint data learning panel summary
- Author
-
Erik Blasch, Andreas Savakis, Yufeng Zheng, Genshe Chen, Ivan Kadar, Uttam Majumder, and Ali K. Raz
- Published
- 2022
9. Toward Near-Real-Time Training With Semi-Random Deep Neural Networks and Tensor-Train Decomposition
- Author
-
Uttam Majumder, Ryan Bryla, Dhireesha Kudithipudi, and Humza Syed
- Subjects
Atmospheric Science ,neural networks with random weights ,QC801-809 ,Computer science ,Random projection ,Geophysics. Cosmic physics ,Training (meteorology) ,synthetic aperture radar (SAR) ,Computational resource ,Convolutional neural networks (CNNs) ,Matrix decomposition ,Ocean engineering ,extreme learning machine (ELM) ,Memory management ,Compression (functional analysis) ,Decomposition (computer science) ,Network performance ,Computers in Earth Sciences ,TC1501-1800 ,Algorithm ,image classification ,random vector functional link (RVFL) - Abstract
In recent years, deep neural networks have shown to achieve state-of-the-art performance on several classification and prediction tasks. However, these networks demand undesirable lengthy training times coupled with high computational resources (memory, I/O, processing time). In this work, we explore semi-random deep neural networks to achieve near real-time training and less computational resource usage. Although many works enhance the underlying hardware for real-time training, this work focuses on algorithmic optimization. It is shown that random projection networks with additional skipped connectivity and randomly weighted layers can boost the overall network performance while enabling for real-time training. Additionally, a tensor-train decomposition technique is leveraged to further reduce the model complexity of these networks. Our investigation accomplishes the following: 1) Tensor-train decomposition decreases the complexity of random projection networks, 2) compression of the fully connected hidden layer leads to a minimum $\sim40\times$ decrease in memory size, and 3) training under random projection networks can be achieved in near-real time.
- Published
- 2021
10. Polyembolokoilamania with obsessive compulsive and related disorders: A case series
- Author
-
AvikKumar Layek, Uttam Majumder, and Iti Baidya
- Subjects
Psychiatry and Mental health - Published
- 2023
11. Diagnostic Variations and Treatment Responses of Catatonic Patients in the Psychiatric In-Patient Department of a Rural Tertiary Care Centre - A Clinical Study from Eastern India
- Author
-
Avik Kumar Layek and Uttam Majumder
- Subjects
Clinical study ,medicine.medical_specialty ,business.industry ,Family medicine ,medicine ,In patient ,business ,Tertiary care ,Eastern india - Published
- 2020
12. Methods of AI for Multimodal Sensing and Action for Complex Situations
- Author
-
Todd V. Rovito, Uttam Majumder, Erik Blasch, Robert Cruise, and Alexander Aved
- Subjects
Computer science ,business.industry ,Swarm behaviour ,Context (language use) ,02 engineering and technology ,Sensor fusion ,Contextual design ,Software ,Action (philosophy) ,Artificial Intelligence ,Human–computer interaction ,Data in use ,0202 electrical engineering, electronic engineering, information engineering ,020201 artificial intelligence & image processing ,Data in transit ,business - Abstract
Artificial intelligence (AI) seeks to emulate human reasoning, but is still far from achieving such results for actionable sensing in complex situations. Instead of emulating human situation understanding, machines can amplify intelligence by accessing large amounts of data, filtering unimportant information, computing relevant context, and prioritizing results (for example, answers to human queries) to provide human–machine shared context. Intelligence support can come from many contextual sources that augment data reasoning through physical, environmental, and social knowledge. We propose a decisions-to-data multimodal sensor and action through contextual agents (human or machine) that seek, combine, and make sense of relevant data. Decisions-to-data combines AI computational capabilities with human reasoning to manage data collections, perform data fusion, and assess complex situations (that is, context reasoning). Five areas of AI developments for context-based AI that cover decisions-to-data include: (1) situation modeling (data at rest), (2) measurement control (data in motion), (3) statistical algorithms (data in collect), (4) software computing (data in transit), and (5) human–machine AI (data in use). A decisions-to-data example is presented of a command-guided swarm requiring contextual data analysis, systems-level design, and user interaction for effective and efficient multimodal sensing and action.
- Published
- 2019
13. Review of recent advances in AI/ML using the MSTAR data
- Author
-
Erik Blasch, Uttam Majumder, Edmund G. Zelnio, and Vincent J. Velten
- Subjects
Synthetic aperture radar ,business.industry ,Computer science ,Deep learning ,ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION ,Context (language use) ,Machine learning ,computer.software_genre ,Target acquisition ,Data set ,Identification (information) ,ComputingMethodologies_PATTERNRECOGNITION ,Automatic target recognition ,Artificial intelligence ,business ,computer - Abstract
Over the past decades, there have been many approaches to synthetic aperture radar (SAR) automatic target recognition (ATR). ATR includes detection, classification, and identification of targets, scene, and context. Recently, the explosion of methods for deep learning has attracted numerous researchers to compare machine learning methods for SAR ATR. This paper reviews many approaches conducted for SAR recognition and discerns the most promising approaches. Using the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set, there are comparative methods to evaluate the advances from the community. The paper reviews many of the available techniques recently published to determine the state of the art in emerging concepts.
- Published
- 2020
14. Sparsity-based data conditioning for more robust deep-learning synthetic aperture radar target classification (Conference Presentation)
- Author
-
Eric K. Davis, Chris Capraro, Francesca Vidal, and Uttam Majumder
- Subjects
Synthetic aperture radar ,Presentation ,Computer science ,business.industry ,Deep learning ,media_common.quotation_subject ,Computer vision ,Artificial intelligence ,business ,media_common - Published
- 2020
15. A method for detecting the presence of confusion targets in synthetic aperture radar target classification (Conference Presentation)
- Author
-
Uttam Majumder, Wesley Stevens, Chris Capraro, and Eric K. Davis
- Subjects
Synthetic aperture radar ,Presentation ,business.industry ,Computer science ,media_common.quotation_subject ,medicine ,Computer vision ,Artificial intelligence ,medicine.symptom ,business ,Confusion ,media_common - Published
- 2020
16. Advanced Techniques for Robust SAR ATR: Mitigating Noise and Phase Errors
- Author
-
Nathan Inkawhich, Chris Capraro, Yi Chen, Eric A. Davis, and Uttam Majumder
- Subjects
Synthetic aperture radar ,Contextual image classification ,Computer science ,business.industry ,Deep learning ,Real-time computing ,0211 other engineering and technologies ,020206 networking & telecommunications ,02 engineering and technology ,Adversarial machine learning ,Automatic target recognition ,Robustness (computer science) ,Phase noise ,0202 electrical engineering, electronic engineering, information engineering ,Artificial intelligence ,business ,Search and rescue ,021101 geological & geomatics engineering - Abstract
We present advanced Deep Learning (DL) techniques for robust Synthetic Aperture Radar (SAR) automatic target recognition (ATR) in the presence of noise and signal phase errors. Our research focuses on ensuring robust performance of SAR ATR algorithms under noise and adversarial attacks. Robust DL-based SAR ATR is paramount in operational scenarios such as disaster relief, search and rescue, and highly accurate object classification for autonomous vehicles. Our contributions, as described in this paper, include algorithm development and implementation of an advanced deep learning technique known as adversarial training (AT) to mitigate the detrimental effects of sophisticated noise and phase errors. Our research demonstrated that 1) AT improves performance under extended operating conditions, in some cases improving up to 10% over models without AT. 2) The use of AT improves performance when sinusoidal or wideband phase noise is present, in some cases gaining 40% in accuracy that would be lost in the presence of noise. 3) We find the model architecture has significant impact on robustness, with more complex networks showing a greater improvement from AT. 4) The availability of multi-polarization data is always advantageous. To our knowledge no one has provided an extensive analysis of the impact of adversarial machine learning (ML) on SAR image classification. Thus, this paper serves as a comprehensive research revealing the impact of adversarial attack and how to mitigate it.
- Published
- 2020
17. Artificial Intelligence in Use by Multimodal Fusion
- Author
-
Erik P. Blasch, Uttam Majumder, Todd Rovito, and Ali K. Raz
- Published
- 2019
18. SAR object classification implementation for embedded platforms
- Author
-
Chris Capraro, Daniel Brown, Uttam Majumder, Eric K. Davis, Chris Cicotta, and Josh Siddall
- Subjects
Synthetic aperture radar ,Artificial neural network ,Computer science ,business.industry ,Computation ,Pattern recognition ,Convolutional neural network ,TrueNorth ,ComputingMethodologies_PATTERNRECOGNITION ,Neuromorphic engineering ,Artificial intelligence ,IBM ,business ,Operation - Abstract
This research details a new approach to optimize neural network architectures for Synthetic Aperture Radar (SAR) object classification on neuromorphic (e.g., IBM’s TrueNorth) and embedded platforms. We developed an algorithm to reduce the run-time and the power consumption of Deep Neural Networks (DNNs) classifiers by reducing the DNN model size required for a given object classification task. Reducing the model size reduces the number of mathematical operations performed, and the memory required, enabling computation on low size, weight and power (SWaP) hardware. We will provide our approach and results on relevant SAR data. Our entirely new approach starts with a very small multi-class convolution neural network (CNN) and replaces the standard negative log likelihood loss function with a single-class log loss function. We then generate an ensemble of small models trained for an individual class by varying the training data using a k-fold cross-validation and augmentation. This is done for each class and the resulting ensembles classify objects by finding the maximum average probability across each ensemble of single-class classifiers. We demonstrate 91-99 percent classification accuracy on three different datasets with composite networks that require almost 10 times fewer mathematical operations than SqueezeNet (a reduced parameter CNN with AlexNet performance).
- Published
- 2019
19. Semi-random deep neural networks for near real-time target classification
- Author
-
Dhireesha Kudithipudi, Uttam Majumder, Ryan Bryla, and Humza Syed
- Subjects
Artificial neural network ,Exploit ,business.industry ,Computer science ,Training (meteorology) ,Image processing ,Machine learning ,computer.software_genre ,Backpropagation ,Fraction (mathematics) ,Network performance ,Artificial intelligence ,business ,computer ,Randomness - Abstract
In recent years deep neural networks have shown great advances in image processing tasks. For modern datasets, these networks require long training times due to backpropagation, high amount of computational resources for weight updates, and memory intensive weight storage. Exploiting randomness during the training of deep neural networks can mitigate these concerns by reducing the computational costs without sacrificing network performance. However, a fully randomized network has limitations for real-time target classification as it leads to poor performance. Therefore we are motivated in using semi-random deep neural networks to exploit random fixed weights. In this paper, we demonstrate that semi-random deep neural networks can achieve near real-time training with comparable accuracies to conventional deep neural networks models. We find that these networks are enhanced by the usage of skip connections and train rapidly at the cost of dense memory usage. With greater memory resources available, these networks can train on larger datasets at a fraction of the training time costs. These semi-random deep neural network architectures open up an avenue for further research in utilizing random fixed weights in neural networks.
- Published
- 2019
20. An approaches for noise induced object classifications accuracy improvement
- Author
-
Chris Capraro, Uttam Majumder, Josh Siddall, Chris Cicotta, Daniel Brown, and Eric K. Davis
- Subjects
Synthetic aperture radar ,Contextual image classification ,business.industry ,Image quality ,Computer science ,ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION ,Cognitive neuroscience of visual object recognition ,Pattern recognition ,Object (computer science) ,Object detection ,Noise ,Compressed sensing ,Computer Science::Computer Vision and Pattern Recognition ,Artificial intelligence ,business - Abstract
Among various parameters, large scene object detection and classification accuracy depends on image quality. In general, deep neural networks (DNN) are trained to achieve a desired recognition accuracy on a set of targets. However, DNNs become tuned to the training data used and may not generalize to new unseen data artifacts. Classification accuracy of a previously trained DNN is significantly reduced when classification is run on an image altered with additive noise. In this research, we propose image pre-processing to reduce the impact of noise induced low classification accuracy. Our approach consists of applying compressive sensing inspired pre-processing techniques to noisy images. We then compare the object recognition accuracy of a pretrained model on pre-processed noisy images and unprocessed noisy images. We will present our technical method, results, and analysis on relevant synthetic aperture radar data.
- Published
- 2019
21. Automatic machine learning for target recognition
- Author
-
Erik Blasch, Todd V. Rovito, Vincent J. Velten, Uttam Majumder, and Peter Zulch
- Subjects
Signal processing ,System deployment ,Computer science ,business.industry ,Multimodal data ,Machine learning ,computer.software_genre ,Information fusion ,Robust learning ,Automatic target recognition ,Robustness (computer science) ,Artificial intelligence ,Transfer of learning ,business ,computer - Abstract
Automatic Target Recognition (ATR) seeks to improve upon techniques from signal processing, pattern recognition (PR), and information fusion. Currently, there is interest to extend traditional ATR methods by employing Artificial Intelligence (AI) and Machine Learning (ML). In support of current opportunities, the paper discusses a methodology entitled: Systems Experimentation efficiency effectives Evaluation Networks (SEeeEN). ATR differs from PR in that ATR is a system deployment leveraging pattern recognition (PR) in a networked environment for mission decision making, while PR/ML is a statistical representation of patterns for classification. ATR analysis has long been part of the COMPrehensive Assessment of Sensor Exploitation (COMPASE) Center utilizing measures of performance (e.g., efficiency) and measures of effectiveness (e.g., robustness) for ATR evaluation. The paper highlights available multimodal data sets for Automated ML Target Recognition (AMLTR).
- Published
- 2019
22. Multisource deep learning for situation awareness
- Author
-
Zheng Liu, Yufeng Zheng, Peter Zulch, Alexandar Aved, Erik Blasch, and Uttam Majumder
- Subjects
Image fusion ,Situation awareness ,Computer science ,business.industry ,Deep learning ,Contextual reasoning ,Machine learning ,computer.software_genre ,Convolutional neural network ,law.invention ,Multimodal learning ,law ,Robustness (computer science) ,Artificial intelligence ,Radar ,business ,computer - Abstract
The resurgence of interest in artificial intelligence (AI) stems from impressive deep learning (DL) performance such as hierarchical supervised training using a Convolutional Neural Network (CNN). Current DL methods should provide contextual reasoning, explainable results, and repeatable understanding that require evaluation methods. This paper discusses DL techniques using multimodal (or multisource) information that extend measures of performance (MOP). Examples of joint multi-modal learning include imagery and text, video and radar, and other common sensor types. Issues with joint multimodal learning challenge many current methods and care is needed to apply machine learning methods. Results from Deep Multimodal Image Fusion (DMIF) using Electro-optical and infrared data demonstrate performance modeling based on distance to better understand DL robustness and quality to provide situation awareness.
- Published
- 2019
23. Deep learning in AI and information fusion panel discussion
- Author
-
Ivan Kadar, Garrett Stevenson, Chee-Yee Chong, Lynne L. Grewe, Uttam Majumder, and Erik Blasch
- Subjects
Multimodal fusion ,Information fusion ,business.industry ,Human–computer interaction ,Computer science ,Deep learning ,Big data ,Cognitive neuroscience of visual object recognition ,Context (language use) ,Cognition ,Artificial intelligence ,business ,Panel discussion - Abstract
During the 2018 SPIE DSS conference, panelists were invited to highlight the trends and use of artificial intelligence and deep learning (AI/DL) for information fusion. This paper highlights the common issues presented from the panel discussion. The key issues include: leveraging AI/DL coordinated with information fusion for: (1) knowledge reasoning and reasoning, (2) information fusion enhancement, (3) object recognition and tracking, (4) data with models fusion, and (5) deep multimodal fusion cognition strategies to support the user.
- Published
- 2019
24. IEEE Radar Conference 2019 Tutorial Session I Machine Learning Techniques for Radar ATR: Monday, April 22, 2019 Time: 11:00AM – 3:00PM
- Author
-
Erik Blasch, David A. Garren, and Uttam Majumder
- Subjects
Engineering ,business.industry ,0211 other engineering and technologies ,02 engineering and technology ,021001 nanoscience & nanotechnology ,TrueNorth ,law.invention ,Aeronautics ,law ,Internship ,Session (computer science) ,Radar ,IBM ,0210 nano-technology ,business ,021101 geological & geomatics engineering - Abstract
Acknowledgements •DARPA MSTAR Program •DARPA TRACE PM (Dr. John Gorman) •DARPA RFMLS PM (Mr. Paul Tilghman) •IBM's TrueNorth Program (Dr. Dharmendra Modha) •AFRL Colleagues (Dr. Qing Wu, Mr. Edmund Zelnio, Dr. Vince Velten, Ms. Lori Westerkamp, Dr. Eric Branch, Dr. John Nehrbass) •AFRL Information Directorate Summer Internship Program •Mr. Nathan Inkawhich (Duke University) •Mr. Matthew Inkawhich (Duke University)
- Published
- 2019
25. Design and analysis of radar waveforms achieving transmit and receive orthogonality
- Author
-
Mark R. Bell, Muralidhar Rangaswamy, and Uttam Majumder
- Subjects
Engineering ,business.industry ,Pulse-Doppler radar ,010401 analytical chemistry ,Aerospace Engineering ,020206 networking & telecommunications ,02 engineering and technology ,Direct-sequence spread spectrum ,01 natural sciences ,0104 chemical sciences ,law.invention ,Continuous-wave radar ,Spread spectrum ,Bistatic radar ,law ,0202 electrical engineering, electronic engineering, information engineering ,Chirp ,Electronic engineering ,Electrical and Electronic Engineering ,Radar ,business ,Diversity scheme - Abstract
This paper presents the design and analysis of orthogonal, Doppler-tolerant waveforms for waveform agile radar (e.g. multiple-input multiple-output (MIMO) radar) applications. Previous work has given little consideration to the design of radar waveforms that remain orthogonal when they are received. Our research is focused on: 1) developing sets of waveforms that are orthogonal on both transmit and receive, and 2) ensuring that these waveforms are Doppler tolerant when properly processed. Our proposed solution achieves the above-mentioned goals by incorporating direct sequence spread spectrum (DSSS) coding techniques on linear frequency modulated (LFM) signals. We call this spread spectrum coded LFM (SSCL) signaling. Our transmitted LFM waveforms are rendered orthogonal with a unique spread spectrum (SS) code. At the receiver, the echo signal will be decoded using its spreading code. In this manner, transmitted orthogonal waveforms can be match filtered only with the intended received signals. From analytical expressions of the waveforms we have designed and from simulation results, we found that: 1) cross-ambiguity function (CAF) of two LFM SS coded (orthogonal) waveforms is small for all delays and Dopplers (i.e. transmit and receive signals satisfy near orthogonality constraint); 2) the length of the SS code determines the amount of interference suppression (i.e., complete orthogonal or near orthogonal of the received signal); 3) we can process the same received signal in two different ways; one method can provide LFM signal resolution and the other method can provide ultrahigh resolution; 4) biorthogonal codes can be used to reduce bandwidth when code length is large. Our proposed waveforms inherit multiple attributes (e.g. chirp diversity, code diversity, frequency diversity) of diverse waveforms.
- Published
- 2016
26. SAR geolocation of moving targets using knowledge of network of roads (Conference Presentation)
- Author
-
Michael J. Minardi, Alexander Boytim, Mehrdad Soumekh, and Uttam Majumder
- Subjects
Geolocation ,Presentation ,Information retrieval ,Computer science ,media_common.quotation_subject ,media_common - Published
- 2018
27. Wavelet decomposition to reduce clutter for SAR object classification using deep neural networks (Conference Presentation)
- Author
-
Nathan Inkawhich and Uttam Majumder
- Subjects
Presentation ,Wavelet decomposition ,Computer science ,business.industry ,media_common.quotation_subject ,Deep neural networks ,Clutter ,Pattern recognition ,Artificial intelligence ,Object (computer science) ,business ,media_common - Published
- 2018
28. A review on image quality and computational costs for various SAR imaging algorithms for machine learning applications (Conference Presentation)
- Author
-
Uttam Majumder and Mehrdad Soumekh
- Subjects
Presentation ,Computer science ,business.industry ,Image quality ,media_common.quotation_subject ,Artificial intelligence ,Machine learning ,computer.software_genre ,business ,computer ,media_common - Published
- 2018
29. Synthetic RF data for advanced machine learning algorithms development (Conference Presentation)
- Author
-
Eric Branch, Uttam Majumder, John Nehrbass, and Edmund G. Zelnio
- Subjects
Presentation ,Development (topology) ,Multimedia ,Computer science ,media_common.quotation_subject ,Transfer of learning ,computer.software_genre ,computer ,media_common - Published
- 2018
30. Cataract Surgery Outcomes in Bangladeshi Children
- Author
-
Johurul Jewel, Riton Deb, Uttam Majumder, Richard Bowman, Clare Gilbert, Tariq Ayoub, Sadeq Ahmed, Mohammad Muhit, and Guy S. Negretti
- Subjects
Adult ,Male ,Rural Population ,medicine.medical_specialty ,Pediatrics ,Visual acuity ,Adolescent ,Urban Population ,genetic structures ,medicine.medical_treatment ,Vision Disorders ,Visual Acuity ,Intraocular lens ,Cataract Extraction ,Blindness ,Cataract ,Young Adult ,Postoperative Complications ,Surveys and Questionnaires ,Ophthalmology ,Humans ,Medicine ,Young adult ,Child ,Intraoperative Complications ,Dioptre ,Bangladesh ,business.industry ,Cataract surgery ,Childhood cataract ,eye diseases ,Child, Preschool ,Cohort ,Female ,medicine.symptom ,Complication ,business - Abstract
To measure visual acuity (VA) outcomes, complication rates, and the social impact of cataract surgery in a cohort who underwent surgery as children in Bangladesh.Case series.A total of 471 of 850 children from 6 Bangladeshi districts who had been identified as cataract blind using key informants (KIs) between 2004 and 2009 during the Bangladesh Childhood Cataract Campaign (BCCC) together with all those children not included in the BCCC database but in the Child Sight Foundation (CSF) database who had been identified as cataract blind.The subjects and families were contacted again by KIs and transported to local examination centers, where parents and subjects were administered a questionnaire and subjects underwent full ocular examination. Where operative data were available (15%), they were analyzed in conjunction with questionnaire and examination findings. Statistical analysis was performed using SPSS Statistics (IBM, Armonk, NY).Presenting and best-corrected visual acuities (BCVAs), cause(s) of poor outcome, postoperative refraction, and school attendance.A total of 407 of the participants had undergone bilateral surgery as children, with a mean follow-up of 8.8 years. The mean age at examination was 16 years (range, 5-28 years; standard deviation [SD], 4.6 years); 63% of those examined were male; 22% had a binocular presenting VA of20/60; and 53% were severely visually impaired or blind (VA20/200). After refraction, 33% had VA20/60 in their better eye and 33% had VA20/200. Factors that predicted poor VA in multivariate logistic regression analysis were nystagmus (P0.001), longer delay in presentation (P0.001), and magnitude of absolute spherical equivalent refractive error (P0.001). Some 50% had nystagmus, and 69% of those currently aged ≤16 years were attending school. Better acuity was associated with school attendance (P0.001), whereas gender was not.Approximately one third of all participants had a BCVA of ≥20/60 in their better eye. Amblyopia and nystagmus limited visual outcome, indicating the need for earlier detection and treatment. This is the first study to show the link between pediatric cataract outcome and access to education, a millennium development goal.
- Published
- 2015
31. Synthetic aperture radar imagery classification using deep neural networks on a neurosynaptic processor (Conference Presentation)
- Author
-
Uttam Majumder
- Subjects
Synthetic aperture radar ,Presentation ,Artificial neural network ,Computer science ,business.industry ,media_common.quotation_subject ,Deep neural networks ,Computer vision ,Artificial intelligence ,business ,Synthetic aperture radar imagery ,media_common - Published
- 2017
32. Machine Learning (ML) Algorithms: An overview of various techniques for target detection and classification (Conference Presentation)
- Author
-
Uttam Majumder
- Subjects
Presentation ,business.industry ,Computer science ,media_common.quotation_subject ,Artificial intelligence ,Machine learning ,computer.software_genre ,business ,computer ,Convolutional neural network ,media_common - Published
- 2017
33. High-performance computing for automatic target recognition in synthetic aperture radar imagery
- Author
-
Erik Blasch, John Nehrbass, Erik Christiansen, Qing Wu, Nathan Inkawhich, and Uttam Majumder
- Subjects
Synthetic aperture radar ,Artificial neural network ,Computer science ,business.industry ,Deep learning ,0211 other engineering and technologies ,02 engineering and technology ,Machine learning ,computer.software_genre ,Supercomputer ,Convolutional neural network ,law.invention ,Data modeling ,Automatic target recognition ,law ,Artificial intelligence ,Radar ,business ,computer ,021101 geological & geomatics engineering - Abstract
Many research efforts have been devoted to applying machine learning (ML) algorithms to the task of Automatic Target Recognition (ATR). In the 90’s, ML techniques such as Neural Networks were less popular due to various technological barriers and applications. Computational resources were scarce and expensive. Today, computational resources are not as expensive as in the past; however, an abundance of sensors and business data need to be analyzed in real-time. High performance computing (HPC) enables ML-based decision making in real-time or near real-time. This research explores the application of deep learning algorithms, specifically convolutional neural networks, to the task of ATR in synthetic aperture radar (SAR) imagery. We developed a Convolution Neural Networks (CNN) architecture for achieving ATR in SAR imagery and found that classification accuracy levels of 99% can be achieved through the application of neural networks. We used graphics processing units (GPU) to accomplish the computational tasks.
- Published
- 2017
34. Agile waveforms for joint SAR-GMTI processing
- Author
-
Jeffrey Corbeil, Stephen McMurray, Mark R. Bell, Steven Jaroszewski, Allan Corbeil, Michael J. Minardi, and Uttam Majumder
- Subjects
Synthetic aperture radar ,020301 aerospace & aeronautics ,Computer science ,ComputerApplications_COMPUTERSINOTHERSYSTEMS ,02 engineering and technology ,Moving target indication ,law.invention ,Spread spectrum ,Continuous-wave radar ,symbols.namesake ,0203 mechanical engineering ,law ,Radar imaging ,symbols ,Electronic engineering ,Chirp ,Waveform ,Radar ,Doppler effect ,Remote sensing - Abstract
Wideband radar waveforms that employ spread-spectrum techniques were investigated and experimentally tested. The waveforms combine bi-phase coding with a traditional LFM chirp and are applicable to joint SAR-GMTI processing. After de-spreading, the received signals can be processed to support simultaneous GMTI and high resolution SAR imaging missions by airborne radars. The spread spectrum coding techniques can provide nearly orthogonal waveforms and offer enhanced operations in some environments by distributing the transmitted energy over a large instantaneous bandwidth. The LFM component offers the desired Doppler tolerance. In this paper, the waveforms are formulated and a shift-register approach for de-spreading the received signals is described. Hardware loop-back testing has shown the feasibility of using these waveforms in experimental radar test bed.
- Published
- 2016
35. Performance evaluation of SAR/GMTI algorithms
- Author
-
David Sobota, Ryan E. McGinnis, Wendy L. Garber, Uttam Majumder, Michael J. Minardi, and William Pierson
- Subjects
Pulse repetition frequency ,Synthetic aperture radar ,020301 aerospace & aeronautics ,Computer science ,Bandwidth (signal processing) ,02 engineering and technology ,Filter (signal processing) ,01 natural sciences ,Moving target indication ,law.invention ,Data modeling ,010309 optics ,symbols.namesake ,0203 mechanical engineering ,law ,0103 physical sciences ,symbols ,Waveform ,Radar ,Algorithm ,Doppler effect - Abstract
There is a history and understanding of exploiting moving targets within ground moving target indicator (GMTI) data, including methods for modeling performance. However, many assumptions valid for GMTI processing are invalid for synthetic aperture radar (SAR) data. For example, traditional GMTI processing assumes targets are exo-clutter and a system that uses a GMTI waveform, i.e. low bandwidth (BW) and low pulse repetition frequency (PRF). Conversely, SAR imagery is typically formed to focus data at zero Doppler and requires high BW and high PRF. Therefore, many of the techniques used in performance estimation of GMTI systems are not valid for SAR data. However, as demonstrated by papers in the recent literature, 1-11 there is interest in exploiting moving targets within SAR data. The techniques employed vary widely, including filter banks to form images at multiple Dopplers, performing smear detection, and attempting to address the issue through waveform design. The above work validates the need for moving target exploitation in SAR data, but it does not represent a theory allowing for the prediction or bounding of performance. This work develops an approach to estimate and/or bound performance for moving target exploitation specific to SAR data. Synthetic SAR data is generated across a range of sensor, environment, and target parameters to test the exploitation algorithms under specific conditions. This provides a design tool allowing radar systems to be tuned for specific moving target exploitation applications. In summary, we derive a set of rules that bound the performance of specific moving target exploitation algorithms under variable operating conditions.
- Published
- 2016
36. The development of advanced spread spectrum LFM waveforms for enhanced SAR and GMTI
- Author
-
Michael J. Minardi, John C. Kirk, Mark R. Bell, Scott Darden, and Uttam Majumder
- Subjects
Synthetic aperture radar ,020301 aerospace & aeronautics ,Signal processing ,Computer science ,Matched filter ,02 engineering and technology ,Direct-sequence spread spectrum ,Moving target indication ,law.invention ,Spread spectrum ,0203 mechanical engineering ,law ,Electronic engineering ,Waveform ,Radar ,Remote sensing - Abstract
Advanced spread spectrum linear frequency modulated (LFM) waveforms are being developed for advanced capability synthetic aperture radar (SAR) and ground moving target indication (GMTI) applications. We have demonstrated by analysis and simulation the feasibility of these new type waveforms and are now in the process of implementing them in hardware. The basic approach is to combine a traditional LFM radar waveform with a direct sequence spread spectrum (DSSS) waveform, and then on receive to de-spread the return and capture the resultant LFM return for traditional matched filter processing and enhanced SAR and GMTI. We show the analysis, simulation and some preliminary hardware results.
- Published
- 2016
37. Ground moving target processing for tracking selected targets
- Author
-
Gregory Owirka, Howard Nichols, Lucas Finn, and Uttam Majumder
- Subjects
020301 aerospace & aeronautics ,Vehicle tracking system ,Radar tracker ,business.industry ,Computer science ,Tracking system ,02 engineering and technology ,Tracking (particle physics) ,law.invention ,0203 mechanical engineering ,law ,Clutter ,Computer vision ,Artificial intelligence ,Radar ,business ,Particle filter - Abstract
BAE Systems has developed a baseline real-time selected vehicle (SV) radar tracking capability that successfully tracked multiple civilian vehicles in real-world traffic conditions within challenging semi-urban clutter. This real-time tracking capability was demonstrated in laboratory setting. Recent enhancements to the baseline capability include multiple detection modes, improvements to the system-level design, and a wide-area tracking mode. The multiple detection modes support two tracking regimes; wide-area and localized selected vehicle tracking. These two tracking regimes have distinct challenges that may be suited to different trackers. Incorporation of a wide-area tracking mode provides both situational awareness and the potential for enhancing SV track initiation. Improvements to the system-level design simplify the integration of multiple detection modes and more realistic SV track initiation capabilities. Improvements are designed to contribute to a comprehensive tracking capability that exploits a continuous stare paradigm. In this paper, focus will be on the challenges, design considerations, and integration of selected vehicle tracking.
- Published
- 2016
38. Statistical performance analysis for GMTI using ATI phase distribution
- Author
-
Uttam Majumder, Ke Yong Li, Unnikrishna Pillai, David Sobota, and Michael J. Minardi
- Subjects
Synthetic aperture radar ,business.industry ,Scattering ,Computer science ,Aperture ,02 engineering and technology ,021001 nanoscience & nanotechnology ,01 natural sciences ,Moving target indication ,010309 optics ,Inverse synthetic aperture radar ,Interferometry ,0103 physical sciences ,Clutter ,High range resolution ,Computer vision ,Artificial intelligence ,False alarm ,0210 nano-technology ,business ,Image resolution - Abstract
Synthetic aperture radar (SAR) imaging is often used to image an area using airborne platforms that generate a large aperture by virtue of the platform motion. Large apertures generate a large synthetic array providing fine cross-range resolution, and together with wide bandwidth waveforms that provide high range resolution, fine resolution images can be generated. SAR algorithms make use of coherent phase compensation from various pulses for focusing and the technique works exceedingly well for scenes containing stationary scattering centers. When moving targets are present, their images are smeared and shifted due to the motion, and to take advantage of this shift, nearby receiver plates are used to form multiple SAR images and together with along track interferometry (ATI), it generates a phase factor that can be used to detect moving target presence. This paper examines the distribution of the phase variable used in ATI for zero mean Complex Gaussian clutter/target data, and uses the results to address the target in clutter problem as a hypothesis testing problem to compute the probability of detection/false alarm as a function of target to clutter ratio and its velocity.
- Published
- 2016
39. Analytical SAR-GMTI principles
- Author
-
Christopher F. Barnes, Mehrdad Soumekh, Michael J. Minardi, Uttam Majumder, and David Sobota
- Subjects
Synthetic aperture radar ,business.industry ,Computer science ,0211 other engineering and technologies ,02 engineering and technology ,Tracking (particle physics) ,Moving target indication ,Signature (logic) ,symbols.namesake ,Coherent processing interval ,symbols ,Clutter ,Computer vision ,Artificial intelligence ,business ,Doppler effect ,021101 geological & geomatics engineering - Abstract
This paper provides analytical principles to relate the signature of a moving target to parameters in a SAR system. Our objective is to establish analytical tools that could predict the shift and smearing of a moving target in a subaperture SAR image. Hence, a user could identify the system parameters such as the coherent processing interval for a subaperture that is suitable to localize the signature of a moving target for detection, tracking and geolocating the moving target. The paper begins by outlining two well-known SAR data collection methods to detect moving targets. One uses a scanning beam in the azimuth domain with a relatively high PRF to separate the moving targets and the stationary background (clutter); this is also known as Doppler Beam Sharpening. The other scheme uses two receivers along the track to null the clutter and, thus, provide GMTI. We also present results on implementing our SAR-GMTI analytical principles for the anticipated shift and smearing of a moving target in a simulated code. The code would provide a tool for the user to change the SAR system and moving target parameters, and predict the properties of a moving target signature in a subaperture SAR image for a scene that is composed of both stationary and moving targets. Hence, the SAR simulation and imaging code could be used to demonstrate the validity and accuracy of the above analytical principles to predict the properties of a moving target signature in a subaperture SAR image.
- Published
- 2016
40. Chronic Burn Ulceration of the Skin and the Potential of Amniotic Membrane-Based Therapy
- Author
-
Diptendu Bikash Sengupta, Sushanta Kumar Banerjee, Biswanath Chakraborty, Prosanto Chowdhury, Uttam Majumder, Kalisankar Das, Dhritidipa Chowdhury, Akash Bhattacharya, Niranjan Bhattacharya, Shyama Prasad Das, Soumen K. Bhattacharya, Dipanjan Sengupta, Tapan Rakshit, Debranjan Basu, Nandita Basu, Samarendranarayan Choudhuri, Raja Bhattacharya, Aditi Aikat, Santanu Kumar Tripathi, Niranjan Maity, Samir Chaudhuri, Sujata Law, Soma Das, Swapna Chaudhuri, and Dhruba Malakar
- Subjects
Chronic wound ,medicine.medical_specialty ,Amniotic fluid ,integumentary system ,business.industry ,Mesenchymal stem cell ,medicine ,medicine.symptom ,business ,Dermatology ,Surgery - Abstract
In a recent report, it has been brought to notice that close to five million people in the USA are affected by chronic wounds and billions are spent annually for their treatment. Despite advances in chronic wound management over the past decades, many patients afflicted with chronic wounds fail to heal or their ulcers recur. Why is this so?
- Published
- 2014
41. Diverse, orthogonal waveforms and signal processing architecture for joint GMTI and SAR applications
- Author
-
Muralidhar Rangaswamy, Uttam Majumder, and Mark R. Bell
- Subjects
Synthetic aperture radar ,Pulse repetition frequency ,Signal processing ,Early-warning radar ,Computer science ,Pulse-Doppler radar ,Side looking airborne radar ,Fire-control radar ,Radar lock-on ,Moving target indication ,law.invention ,Continuous-wave radar ,Inverse synthetic aperture radar ,symbols.namesake ,Space-time adaptive processing ,Bistatic radar ,Radar engineering details ,law ,Radar imaging ,symbols ,Electronic engineering ,Waveform ,Radar ,Doppler effect ,Radar horizon - Abstract
In this research, we introduce a signal processing framework for joint GMTI and SAR algorithms that is based on orthogonal (transmit and receive) waveforms. Traditionally, radar systems are configured to operate either in GMTI or SAR processing mode, but not both simultaneously. This is due to the fact that operational parameters for these two modes are quite different. For example, exoclutter GMTI processing requires a high pulse repetition frequency (PRF), but a high PRF results in increased range ambiguity - and an increased processing burden - in SAR imaging. We propose combining diverse, orthogonal waveforms and introducing corresponding processing techniques to reduce the problems and complexities of joint GMTI and SAR exploitation. For the exoclutter GMTI problem, the necessary high-PRF pulse train will be used to achieve finer Doppler resolution for detecting fast moving objects. For the endoclutter GMTI and SAR imaging problem, we will transmit low PRF pulses. The goal for low PRF pulses for endoclutter GMTI and SAR imaging is to ensure that range ambiguity issue has been addressed. These new approaches will achieve following benefits: (1) accomplish GMTI and SAR processing concurrently by eliminating the complexities associated with reconfiguring a radar system, (2) more efficiently use bandwidth by employing appropriate bandwidth for exoclutter GMTI pulses and SAR image formation pulses, and (3) reduce range ambiguity issue associated with high PRF operation.
- Published
- 2014
42. Forty years of digital SAR and slow GMTI technology
- Author
-
John C. Kirk, Scott Darden, Uttam Majumder, and Steven Scarborough
- Subjects
Synthetic aperture radar ,Radar tracker ,Early-warning radar ,Computer science ,Pulse-Doppler radar ,Fire-control radar ,Side looking airborne radar ,Radar lock-on ,Moving target indication ,law.invention ,Continuous-wave radar ,Inverse synthetic aperture radar ,Man-portable radar ,Bistatic radar ,Radar engineering details ,law ,Radar imaging ,3D radar ,Radar ,Radar configurations and types ,Radar MASINT ,Remote sensing - Abstract
40-years of digital SAR and Slow GMTI technology is traced from the first system to demonstrate real-time digitally correlated SAR from an intentionally maneuvering platform [1] to the current lite-weight dual-channel radar (DCR) providing simultaneous SAR and GMTI data. The Dual-Channel Radar (DCR) has been developed providing lite-weight SAR GMTI capability for Small UAVs. The prototype radar weighs 5-lbs and has demonstrated the extraction of ground moving targets (GMTs) embedded in highresolution SAR imagery data. Sum and difference channel data is used in a DPCA algorithm to extract the GMTs and display them on the Sum channel high resolution SAR image. Heretofore this type of capability has been reserved for much larger systems such as the JSTARS. Previously small liteweight SARs featured only a single channel and only displayed SAR imagery. With the advent of this new capability, SAR GMTI performance is now possible for small UAV class radars for DoD and DHS applications. The DCR is the culmination of multiple Phase II and Phase II plus SBIR efforts over the past 10-years, since 2002, for the Army, DARPA and AFRL.
- Published
- 2014
43. Lightweight SAR GMTI radar technology development
- Author
-
Kai Lin, Winston Kwong, Steven Scarborough, Scott Darden, A. D. Gray, John C. Kirk, Chung Hseih, and Uttam Majumder
- Subjects
Synthetic aperture radar ,Early-warning radar ,Computer science ,Fire-control radar ,Radar lock-on ,Moving target indication ,law.invention ,Inverse synthetic aperture radar ,Continuous-wave radar ,Bistatic radar ,Man-portable radar ,Radar engineering details ,law ,Radar imaging ,3D radar ,Radar ,Remote sensing - Abstract
A small and lightweight dual-channel radar has been developed for SAR data collections. Using standard Displaced Phase Center Antenna (DPCA) radar digital signal processing, SAR GMTI images have been obtained. The prototype radar weighs 5-lbs and has demonstrated the extraction of ground moving targets (GMTs) embedded in high-resolution SAR imagery data. Heretofore this type of capability has been reserved for much larger systems such as the JSTARS. Previously, small lightweight SARs featured only a single channel and only displayed SAR imagery. Now, with the advent of this new capability, SAR GMTI performance is now possible for small UAV class radars.
- Published
- 2013
44. A novel approach for designing diversity radar waveforms that are orthogonal on both transmit and receive
- Author
-
Muralidhar Rangaswamy, Mark R. Bell, and Uttam Majumder
- Subjects
Pulse repetition frequency ,Computer science ,Pulse-Doppler radar ,Code division multiple access ,Doppler radar ,Chirp spread spectrum ,Direct-sequence spread spectrum ,law.invention ,Continuous-wave radar ,Spread spectrum ,symbols.namesake ,law ,Electronic engineering ,symbols ,Waveform ,Radar ,Doppler effect ,Computer Science::Information Theory - Abstract
In this paper, we present an approach to the design of orthogonal, Doppler tolerant waveforms for diversity waveform radar (e.g. MIMO radar). Previous work has given little consideration to the design of radar waveforms that remain orthogonal when they are received. Our research is focused on: (1) developing sets of waveforms that are orthogonal on both transmit and receive, and (2) ensuring that these waveforms are Doppler tolerant when properly processed. Our proposed solution achieves the above mentioned goals by incorporating direct sequence spread spectrum (DSSS) coding techniques on linear frequency modulated (LFM) signals. We call it Spread Spectrum Coded LFM (SSCL) signaling. Our transmitted LFM waveforms are rendered orthogonal with a unique spread spectrum code. At the receiver, the echo signal will be decoded using it's spreading code. In this manner, transmitted orthogonal waveforms can be match filtered only with the intended received signals. From analytical expressions of the waveforms we have designed and from simulation results, we found that: (a) cross-ambiguity function of two LFM spread spectrum coded (orthogonal) waveforms is small for all delays and Dopplers (i.e. transmit and receive signals satisfy the orthogonality constraint), (b) The length and type of the spread spectrum code determines amount of suppression (i.e. complete orthogonal or near orthogonal of the received signal), (c) We can process the same received signal in two different ways; one method can provide LFM signal resolution and the other method can provide ultra high resolution.
- Published
- 2013
45. Dual-channel radar for small UAVs
- Author
-
Uttam Majumder, Winston Kwong, John C. Kirk, Steven Scarborough, Chung Hseih, Kai Lin, Linda J. Moore, Scott Darden, and Andrew Gray
- Subjects
Synthetic aperture radar ,Inverse synthetic aperture radar ,Bistatic radar ,Radar tracker ,Radar engineering details ,Computer science ,Radar imaging ,3D radar ,Space-based radar ,Remote sensing - Abstract
A Dual-Channel Radar (DCR) has been developed providing lite-weight SAR GMTI capability for Small UAVs. The prototype radar weighs 5-lbs and has demonstrated the extraction of ground moving targets (GMTs) embedded in high-resolution SAR imagery data. Sum and difference channel data is used in a DPCA algorithm to extract the GMTs and display them on the Sum channel high resolution SAR image. Heretofore this type of capability has been reserved for much larger systems such as the JSTARS. Previously small liteweight SARs featured only a single channel and only displayed SAR imagery. Now, with the advent of this new capability, SAR GMTI performance is now possible for small UAV class radars.
- Published
- 2012
46. Parallel processing techniques for the processing of synthetic aperture radar data on GPUs
- Author
-
Sartaj Sahni, Sanjay Ranka, Mark S. Schmalz, Uttam Majumder, Bracy Elton, Linda J. Moore, and William Chapman
- Subjects
Synthetic aperture radar ,Pixel ,Parallel processing (DSP implementation) ,business.industry ,Computer science ,Pipeline (computing) ,Radar imaging ,Stratix ,Field-programmable gate array ,business ,Computer hardware ,Auxiliary memory - Abstract
This paper presents a design for the parallel processing of synthetic aperture radar data using one or more Field Programmable Gate Arrays (FPGAs). Our design supports real-time computation of a two-dimensional image from a matrix of echo pulses and their corresponding response values. Components of this design include: (a) central processing pipeline to perform back projection calculations, (b) pre-fetch cache to minimize external memory access latency, (c) memory bridge that serves as the primary on-chip storage for pulse data, and (d) a pixel queue to direct image data in and out of the pipeline. Design parameters may be adjusted to achieve optimum performance, and multiple instances of this design may be replicated on-chip to achieve prespecified performance objectives. We provide a complexity analysis as a function of the input and output parameters. Simulation results based on an implementation of this design show that our design achieves 160 GFLOPs per instance on a simulated Altera Stratix III EP3SL150 FPGA, and scales well for output image size ranging from 500 × 500 pixels to 5,000 × 5,000 pixels.
- Published
- 2011
47. Staring RF signal processing challenges
- Author
-
Uttam Majumder, LeRoy A. Gorham, Linda J. Moore, Steven Scarborough, Jason T. Parker, and Michael J. Minardi
- Subjects
Synthetic aperture radar ,Pulse-Doppler radar ,Computer science ,business.industry ,Moving target indication ,law.invention ,Inverse synthetic aperture radar ,Space-time adaptive processing ,Staring ,law ,Radar imaging ,Computer vision ,Artificial intelligence ,Radar ,business - Abstract
Traditionally, distinct radar modes have been employed to accomplish specific tasks such as imaging an area of interest, or detecting and tracking moving targets. Staring circular synthetic aperture radar (S-CSAR) provides unique opportunities for exploitation of radio frequency (RF) data collected over a large ground spot. The same phase history may be processed in different manners to generate simultaneous S-CSAR products such as 2-D Video SAR, coherent and non-coherent change detection (CCD and NCD), and ground moving target indication (GMTI). Advanced signal processing techniques can take advantage of the S-CSAR geometry to produce 3-D scene reconstructions. The ability to transmit, record and process large volumes of S-CSAR data, to create high fidelity exploitation products, in real-time, poses significant challenges. This paper addresses several open problems in this research area.
- Published
- 2011
48. Synthetic aperture radar data visualization on the iPod Touch
- Author
-
Rhonda J. Vickery, Tracy Burchett, Aaron Fouts, Troy Klein, Michael J. Minardi, and Uttam Majumder
- Subjects
Synthetic aperture radar ,Geographic information system ,business.industry ,Infrared ,Computer science ,Interface (computing) ,ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION ,Visualization ,Data visualization ,Human–computer interaction ,Georeference ,Computer vision ,Artificial intelligence ,Zoom ,business ,Mobile device - Abstract
A major area of focus for the Air Force is sensor performance in urban environments. Aircraft with multiple sensor modalities, such as Synthetic Aperture RADAR (SAR), Infrared (IR), and Electro-Optics (EO), are essential for intelligence, surveillance, and reconnaissance (ISR) of current and future urban battlefields. Although applications exist for visualization of these types of imagery, they usually require at least a laptop computer and internet connection. Field operatives need to be able to access georeferenced information about imagery as part of a Geographic Information System (GIS) on mobile devices. The iPod/iPhone has a 640x480 resolution multi-touch display, making it an excellent device for interacting with georeferenced imagery. We created an iPhone application that loads SAR imagery and allows the user to interact with it. The user multi-touch interface provides pan and zoom capabilities as well as options to change parameters relating to the query. We describe how operatives in the field can use this application to investigate SAR and GIS related problems on the iPhone mobile device, which otherwise would require a computer and Internet connection.
- Published
- 2010
49. A challenge problem for SAR change detection and data compression
- Author
-
Matthew G. Judge, Alan Pieramico, Steven Jaroszewksi, Leslie M. Novak, LeRoy A. Gorham, Linda J. Moore, Steven Scarborough, Laura Spoldi, Michael J. Minardi, and Uttam Majumder
- Subjects
Synthetic aperture radar ,business.industry ,Image quality ,Computer science ,X band ,law.invention ,Compressed sensing ,law ,Radar imaging ,Computer vision ,Artificial intelligence ,Radar ,business ,Change detection ,Data compression ,Remote sensing - Abstract
This document describes a challenge problem whose scope is two-fold. The first aspect is to develop SAR CCD algorithms that are applicable for X-band SAR imagery collected in an urban environment. The second aspect relates to effective data compression of these complex SAR images, where quality SAR CCD is the metric of performance. A set of X-band SAR imagery is being provided to support this development. To focus research onto specific areas of interest to AFRL, a number of challenge problems are defined. The data provided is complex SAR imagery from an AFRL airborne X-band SAR sensor. Some key features of this data set are: 10 repeat passes, single phase center, and single polarization (HH). In the scene observed, there are multiple buildings, vehicles, and trees. Note that the imagery has been coherently aligned to a single reference.
- Published
- 2010
50. An analytical expression for the three-dimensional response of a point scatterer for circular synthetic aperture radar
- Author
-
Uttam Majumder and Linda J. Moore
- Subjects
Physics ,Azimuth ,Synthetic aperture radar ,Optics ,business.industry ,Aperture ,Radar imaging ,Point (geometry) ,business ,Point target ,Image restoration ,Impulse response - Abstract
Three-dimensional (3-D) spotlight-mode synthetic aperture radar (SAR) images of point scatterers provide insight into the achievable effectiveness of exploitation algorithms given a variety of operating parameters such as elevation angle, azimuth or synthetic aperture extent, and frequency bandwidth. Circular SAR, using 360 degrees of azimuth, offers the benefit of persistent surveillance and the potential for 3-D image reconstruction improvement compared with limited aperture SAR due in part to the increase in favorable viewing angles of unknown objects. The response of a point scatter at the origin, or center of the imaging scene, is known and has been quantified for circular SAR in prior literature by a closed-form solution. The behavior of a point scatterer radially displaced from the origin has been previously characterized for circular SAR through implementation of backprojection image reconstructions. Here, we derive a closed-form expression for the response of an arbitrarily located point scatterer given a circular flight path. In addition, the behavior of the response of an off-center point target is compared to that of a point scatterer at the origin. Symmetries within the 3-D point spread functions (PSFs), or impulse response functions (IPRs), are also noted to provide knowledge of the minimum subset of SAR images required to fully characterize the response of a particular point scatterer. Understanding of simple scattering behavior can provide insight into the response of more complex targets, given that complicated targets may sometimes be modeled as an arrangement of geometrically simple scattering objects.
- Published
- 2010
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.