We present a progressive meta-algorithm for single image super resolution of very large and seamless images. Deep single image super resolution backbone networks are capable of upsampling local image patches but are susceptible to severe tiling artifacts when attempting to integrate the patches together into a single large image. For large images such as 4K images, the individual low-resolution patches also comprise a narrow receptive field that lacks context regarding the pixel information from the surrounding patches. Our Progressive meta-algorithm is inspired by prior works in progressive GANs and is designed to resolve inter-patch tiling artifact across varying scales while further incorporating context by expanding the receptive field to include the surrounding patches through multiple rounds of progressive upsampling. Our meta-algorithm is compatible with standard super-resolving backbones and enables super-resolution of very large images such as 4K images while overcoming practical GPU memory limitations with commodity graphics cards. We evaluate our Meta-Algorithm with the challenging task of 16x super resolution for 4096×4096 images using the ESRGAN super-resolving backbone and demonstrate significantly improved image quality versus a baseline patch-based approach as evaluated using the Learned Perceptual Image Patch Similarity (LPIPS) metric.

Learning High-Resolution representations is essential for semantic segmentation. Convolutional neural network (CNN) architectures with downstream and upstream propagation flow are popular for segmentation in medical diagnosis. However, due to performing spatial downsampling and upsampling in multiple stages, information loss is inexorable. On the contrary, connecting layers densely on high spatial resolution is computationally expensive. In this work, we devise a Loose Dense Connection Strategy to connect neurons in subsequent layers with reduced parameters. On top of that, using a m-way Tree structure for feature propagation we propose Receptive Field Chain Network (RFC-Net) that learns high-resolution global features on a compressed computational space. Our experiments demonstrates that RFC Net achieves state-of-the-art performance on Kvasir and CVC-ClinicDB benchmarks for Polyp segmentation. Our code is publicly available at github.com/sourajitcs/RFC-NetAAAI23.

COVID-19 is an air-borne viral infection, which infects the respiratory system in the human body, and it became a global pandemic in early March 2020. The damage caused by the COVID-19 disease in a human lung region can be identified using Computed Tomography (CT) scans. We present a novel approach in classifying COVID-19 infection and normal patients using a Random Forest (RF) model to train on a combination of Deep Learning (DL) features and Radiomic texture features extracted from CT scans of patient’s lungs. We developed and trained DL models using CNN architectures for extracting DL features. The Radiomic texture features are calculated using CT scans and its associated infection masks. In this work, we claim that the RFs classification using the DL features in conjunction with Radiomic texture features enhances prediction performance. The experiment results show that our proposed models achieve a higher True Positive rate with the average Area Under the Receiver Curve (AUC) of 0.9768, 95% Confidence Interval (CI) [0.9757, 0.9780].

COVID-19 is a novel infectious disease responsible for over 1.2 million deaths worldwide as of November 2020. The need for rapid testing is a high priority and alternative testing strategies including x-ray image classification are a promising area of research. However, at present, public datasets for COVID-19 x-ray images have low data volumes, making it challenging to develop accurate image classifiers. Several recent papers have made use of Generative Adversarial Networks (GANs) in order to increase the training data volumes. But realistic synthetic COVID-19 x-rays remain challenging to generate. We present a novel Mean Teacher + Transfer GAN (MTT-GAN) that generates COVID-19 chest x-ray images of high quality. In order to create a more accurate GAN, we employ transfer learning from the Kaggle pneumonia x-ray dataset, a highly relevant data source orders of magnitude larger than public COVID-19 datasets. Furthermore, we employ the Mean Teacher algorithm as a constraint to improve stability of training. Our qualitative analysis shows that the MTT-GAN generates x-ray images that are greatly superior to a baseline GAN and visually comparable to real x-rays. Although board-certified radiologists can distinguish MTT-GAN fakes from real COVID-19 x-rays, quantitative analysis shows that MTT-GAN greatly improves the accuracy of both a binary COVID-19 classifier as well as a multi-class pneumonia classifier as compared to a baseline GAN. Our classification accuracy is favorable as compared to recently reported results in the literature for similar binary and multi-class COVID-19 screening tasks.

We have adapted the MODIS Observation Coverage (Obscov) algorithm to the gridding of the AIRS L1b calibrated brightness temperatures, and developed a “Numeric-Obscov” implementation capable of utilizing the AIRS Tophat PSF dataset [1, 2]. We show that Obscov significantly reduces AIRS gridded-artifacts from a day to a season over winter 2005. The Obscov algorithm provides a physically accurate gridding model using spatial response information within the IFOV. Previously, Obscov has been used with an approximate triangular PSF, to allow for billions of integrals to be calculated per daily grid. We evaluated the Obscov algorithm via inter-comparison with MODIS-Aqua. Over a day, RMS noise is improved at high latitudes at 0.5°×1°. When averaged over a week or more, RMS noise is reduced globally by almost 40%.

Aqua provides a direct measurement of the surface infrared radiation for a near decade (2003-2011) from a single satellite platform with two completely independent instruments (AIRS and MODIS). We produced a 0.5°×1.0° lat-lon daily gridded near fundamental decadal data record (FDDR) of AIRS all sky brightness temperatures (BT) for all 2378 AIRS channels including the 4μ and 12μ surface windows for the 9-year period. We show (i) that decadal difference trends between the spectrally convolved AIRS and MODIS IR surface channels 3.96μ, 4.05μ and 12.03μ are as small as 0.01K, 0.016K and 0.001K per year respectively, suitable to detect year to year decadal trends of AIRS global surface BTs as small as 0.1K over the decade. (ii) The analysis of gridded AIRS surface BT channel 4.16μ and 12.18μ reveals no global annual warming trend between 2003-2011 and negligible trends for both hemispheres in the 9-year period. (iii) Both AIRS channels 4.16μ and 12.18μ show a statistically significant global annual BT warming trend of 0.6K and 0.87K respectively in the Arctic (60N-90N) with yearly changes varying by as much as 1.6K. (iv) Finally, we show that AIRS global surface BT spectral channel 4.16μ has a correlation with annual, seasonal and monthly GISS global station surface temperature data records 0.96, 0.9, 0.82 respectively. We conjecture that direct satellite level 1B AIRS IR radiance sounding observations constitute a FDDR and that well calibrated IR surface BTs can serve as a surrogate for directly observing global climate surface temperatures.

The Aqua spacecraft, launched on May 4, 2002, carries two well-calibrated independent infrared (IR) grating spectrometers Atmospheric Infrared Sounder (AIRS) and Moderate Resolution Imaging Spectrometer (MODIS), which have been continuously returning upwelling IR spectral radiance measurements for over five years. Based on an Aqua Sr. Project Review, estimates of available flight fuel, power, and orbital projections assess the life span of the Aqua satellite, and these two instruments, to be reliable to 2013. Since launch, these instruments have generated petabytes of data, which are managed and made available by the Goddard Space Flight Center (GSFC) Earth Science Data and Information Services Center and GSFC MODAPS. Agencies such as NOAA, DOD, EPA, and USGS use the AIRS data mostly for weather-related applications, whereas MODIS data are used, in addition to some climate-related studies, for studies of weather, oceans, and land processes, aerosols, natural and man-made disasters, and earth ecology. The Science Investigator-led Processing Systems (SIPS) teams have made many of the desired products derived from these data sets available either as level 2 products and/or level 3 gridded product fields. However, no gridded level 3 data products of radiances, either averaged for a grid element, max, min, or as brightness temperatures (BTs), are provided directly by the SIPS. Thus, one impediment that the general community faces in accessing these MODIS produced petabytes of data is storing such large data sets, interpreting the multiformatted data, and transforming it into helpful data sets for climate-research needs. The Service-Oriented Atmospheric Radiance (SOAR) system has been designed to bridge these gaps and overcome the challenges of bringing this rich data source to the science community, by delivering applications that process these valuable radiance data into standard spatial–temporal grids as well as user-defined criteria on demand. SOAR can serve this community with aggregated, enriched, and thinned gridded data sets provided with access to the data on demand, with query and subsetting capabilities across many dimensions. In addition, SOAR provides online user-specified visualization and analysis requests, all accessible via a Web browser. The utility of SOAR is exposed via Web-service routines, using the Simple Object Access Protocol. The Web-service library and supporting technologies (Axis, PostgreSQL, and Tomcat) reside on a University of Maryland Baltimore Campus client server, which interfaces to and invokes algorithms on the process server, a high-performance computer cluster and storage system. These servers are connected to the sensor data stores at the GSFC via a high-speed fiber-optic network connection [10 Gb/s], providing reliable and fast on-demand access to a vast online library of AIRS and current monthly MODIS source data.

Launched on May 4, 2002, Aqua carries two well calibrated independent infrared (IR) instruments, Atmospheric Infrared Sounder (AIRS) and Moderate Resolution Imaging Spectro- radiometer (MODIS), which together could provide a precise decadal measurement across the IR electro- magnetic spectrum from 3.7 um to 15um to form a fundamental climate data record. According to the National Academy of Science - Nation Research Council report,the notion of a fundamental climate data record (FCDR) is defined as "a time series of measurements of sufficient length and continuity to determine climate variability and change from sensor data (e.g., calibrated radiances, brightness temperatures (BT), radar backscatter) that have been improved and quality controlled over time." AIRS and MODIS have been continuously returning upwelling IR spectral radiance measurements while maintaining a nominal grid precision of ~0.2 - 0.4 K BT [2] with quality controlled calibration improvements for more than five years. Based on the Aqua senior Project Review of available flight fuel, power and orbital maneuvers, the assessed life span of the satellite Aqua is estimated to be 2013. AIRS and MODIS instrument teams also expect their passive radiometers to continue providing self-calibrating measurements with the same precision through the year 2013. With these assumptions, we show in this paper that with these two independent sensors on the same platform with on orbit relative validations, this data record would be of sufficient length and continuity to meet the FCDR definition to observe and determine the decadal variability of seasonal to interannual climate processes. In addition, NPOESS preparatory project will launch satellite systems carrying instruments similar to AIRS and MODIS which will enable production of an FCDR record for additional decades.

Atmospheric temperature and moisture profile sounding data spanning more than three decades comprise one of the longest continuous US polar orbiting satellite data records available today. NASA's Aqua satellite launched on May 4, 2002 carries two well calibrated independent IR grating spectrometers, AIRS and MODIS. Both AIRS and MODIS have been used for extensive research and are truly flagship instruments for the Earth Science community, but neither offers an operational product with gridded calibrated radiances. Radiance data are very useful and have been used for operational weather forecasting models, and to determine observed global trends in temperature, ozone, albedo, desertification, aerosols, etc. The radiances are only available in their native orbital swath form at full resolution that can be awkward and bulky to work with. Their co-location on Aqua provides a unique opportunity for common observation of the same air mass with two different instruments over a long time period.