•An Attention U-Net is used to detect deforestation in Sentinel-2 satellite imagery.•Four state-of-the-art deep learning models are compared to Attention U-Net.•Attention U-Net has the smallest ...parameter space and training time.•Attention U-Net produces more accurate deforestation masks than other methods tested.•The performance improvement is greater when using more complex datasets.
In this paper, we implement and analyse an Attention U-Net deep network for semantic segmentation using Sentinel-2 satellite sensor imagery, for the purpose of detecting deforestation within two forest biomes in South America, the Amazon Rainforest and the Atlantic Forest. The performance of Attention U-Net is compared with U-Net, Residual U-Net, ResNet50-SegNet and FCN32-VGG16 across three different datasets (three-band Amazon, four-band Amazon and Atlantic Forest). Results indicate that Attention U-Net provides the best deforestation masks when tested on each dataset, achieving average pixel-wise F1-scores of 0.9550, 0.9769 and 0.9461 for each dataset, respectively. Mask reproductions from each classifier were also analysed, showing that compared to the ground reference Attention U-Net could detect non-forest polygons more accurately than U-Net and overall it provides the most accurate segmentation of forest/deforest compared with benchmark approaches despite its reduced complexity and training time, thus being the first application of an Attention U-Net to an important deforestation segmentation task. This paper concludes with a brief discussion on the ability of the attention mechanism to offset the reduced complexity of Attention U-Net, as well as ideas for further research into optimising the architecture and applying attention mechanisms into other architectures for deforestation detection. Our code is available at https://github.com/davej23/attention-mechanism-unet.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
•We are the first to propose a deep-learning model to predict software reliability.•The deep-learning neural network model has better prediction performance than other models.•The deep-learning model ...has better robustness in software reliability prediction.
Different software reliability models, such as parameter and non-parameter models, have been developed in the past four decades to assess software reliability in the software testing process. Although these models can effectively assess software reliability in certain testing scenarios, no single model can accurately predict the fault number in software in all testing conditions. In particular, modern software is developed with more sizes and functions, and assessing software reliability is a remarkably difficult task. The recently developed deep learning model, called deep neural network (NN) model, has suitable prediction performance. This deep learning model not only deepens the layer levels but can also adapt to capture the training characteristics. A comprehensive, in-depth study and feature excavation ultimately shows the model can have suitable prediction performance. This study utilizes a deep learning model based on the recurrent NN (RNN) encoder–decoder to predict the number of faults in software and assess software reliability. Experimental results show that the proposed model has better prediction performance compared with other parameter and NN models.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK, ZRSKP
Abstract
Observations of astrophysical objects such as galaxies are limited by various sources of random and systematic noise from the sky background, the optical system of the telescope and the ...detector used to record the data. Conventional deconvolution techniques are limited in their ability to recover features in imaging data by the Shannon–Nyquist sampling theorem. Here, we train a generative adversarial network (GAN) on a sample of 4550 images of nearby galaxies at 0.01 < z < 0.02 from the Sloan Digital Sky Survey and conduct 10× cross-validation to evaluate the results. We present a method using a GAN trained on galaxy images that can recover features from artificially degraded images with worse seeing and higher noise than the original with a performance that far exceeds simple deconvolution. The ability to better recover detailed features such as galaxy morphology from low signal to noise and low angular resolution imaging data significantly increases our ability to study existing data sets of astrophysical objects as well as future observations with observatories such as the Large Synoptic Sky Telescope (LSST) and the Hubble and James Webb space telescopes.
High-throughput single cell imaging is a critical enabling and driving technology in molecular and cellular biology, biotechnology, medicine and related areas. Imaging flow cytometry combines the ...single-cell imaging capabilities of microscopy with the high-throughput capabilities of conventional flow cytometry. Recent advances in imaging flow cytometry are remarkably revolutionizing single-cell analysis. This article describes recent imaging flow cytometry technologies and their challenges.
Imaging flow cytometry combines the single-cell imaging capabilities of microscopy with the high-throughput capabilities of conventional flow cytometry. This article describes recent imaging flow cytometry technologies and their challenges.
Emerging evidence suggests that dysregulation of circular RNAs (circRNAs) closely associated with cancer progression. In this paper, we focus on exploring the functional role of circ-LDLRAD3 in ...pancreatic cancer. Gene expression was determined using quantitative reverse transcriptase polymerase chain reaction and Western blot. Cell count kit-8 and 5-ethynyl-2′-deoxyuridine assay were applied to evaluate the proliferation of PANC-1 and SW1990 cells. The migration and invasion of PANC-1 and SW1990 cells were assessed using wound healing assay and transwell invasion assay. Luciferase reporter assay was performed for target validation. The results showed that circ-LDLRAD3 was overexpressed in pancreatic cancer tissues and cell lines. Increased expression of circ-LDLRAD3 was indicative of a poor prognosis in patients with pancreatic cancer. Knockdown of circ-LDLRAD3 repressed the growth of pancreatic cancer in vitro and in vivo. miR-137-3p was identified as a direct target of circ-LDLRAD3. More importantly, upregulation of circ-LDLRAD3 could mitigate the inhibitory effect of miR-137-3p on the proliferation, migration and invasion of pancreatic cancer cells. Besides, circ-LDLRAD3 could regulate the expression of pleiotrophin (PTN) through miR-137-3p. Taken together, knockdown of circ-LDLRAD3 repressed the proliferation, migration and invasion of pancreatic cancer cells through miR-137-3p/PTN axis, providing a new mechanism for pancreatic cancer progression.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
Spray forming (SF) is a newly developed rapid solidification and near-net-shape technique for preparing high-speed steels (HSSs) in recent years. However, a large amount of overspray powders will be ...produced in the SF process, which brings about a serious material pollution and waste. To solve this problem, a new method of the two-step sintering and subsequent heat treatment was used to fabricate high-performance W6Mo5Cr4V2Co5Nb (825 K) HSSs using the overspray powders. The microstructural evolution and mechanical properties by the heat treatment were systematically studied. Results show that near full dense 825 K alloys with many uniformly distributed carbides are obtained by the two-step sintering. After quenching and triple tempering, the alloys consist of the martensite, M6C and MC carbides, and a small amount of retained austenite. The carbides have smaller particle size of less than 3 μm and greater sphericity than the as-sintered alloy. The diffusion of alloying elements determines the dissolution and re-precipitation of carbides. The fabricated 825 K alloy quenched at 1180 °C and triple tempered at 540 °C possesses the best mechanical properties, with a hardness of 64.2 HRC and bending strength of 2858 MPa, which is much higher than that of casting alloy with similar composition. This work provides a feasible way to recycle the overspray powders.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
•Overall accuracy of crop classification reaches 85 %–90 % by using full year UAVSAR.•Polarimetric parameters contribute more than linear polarizations to crop mapping.•The CP parameters are much ...more important than the FD parameters for crop mapping.•The combined use of four acquisitions is adequate to achieve a nearly optimal accuracy.
Accurate and timely information on the distribution of crop types is vital to agricultural management, ecosystem services valuation and food security assessment. Synthetic Aperture Radar (SAR) systems have become increasingly popular in the field of crop monitoring and classification. However, the potential of time-series polarimetric SAR data has not been explored extensively, with several open scientific questions (e.g. the optimal combination of image dates for crop classification) that need to be answered. In this research, the usefulness of full year (both 2011 and 2014) L-band fully-polarimetric Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) data in crop classification was fully investigated over an agricultural region with a heterogeneous distribution of crop categories. In total, 11 crop classes including tree crops (almond and walnut), forage crops (grass, alfalfa, hay, and clover), a spring crop (winter wheat), and summer crops (corn, sunflower, tomato, and pepper), were discriminated using the Random Forest (RF) algorithm. The SAR input variables included raw linear polarization channels as well as polarimetric parameters derived from Cloude-Pottier (CP) and Freeman-Durden (FD) decompositions. Results showed clearly that the polarimetric parameters yielded much higher classification accuracies than linear polarizations. The combined use of all variables (linear polarizations and polarimetric parameters) produced the maximum overall accuracy of 90.50 % and 84.93 % for 2011 and 2014, respectively, with a significant increase of approximately 8 percentage points compared with linear polarizations alone. The variable importance provided by the RF illustrated that the polarimetric parameters had a far greater influence than linear polarizations, with the CP parameters being much more important than the FD parameters. The most important acquisitions were the images dated during the peak biomass stage (July and August) when the differences in structural characteristics between most crops were the largest. At the same time, the images in spring (April and May) and autumn (October) also contributed to the crop classification since they respectively provided unique information for discriminating fruit crops (almond and walnut) as well as summer crops (corn, sunflower, and tomato). As a result, the combined use of only four acquisitions (dated May, July, August, and October for 2011 and April, June, August, and October for 2014) was adequate to achieve a nearly-optimal overall accuracy. In light of the promising classification accuracies demonstrated in this research, it becomes increasingly viable to provide accurate and up-to-date crops inventories over large areas based solely on multitemporal polarimetric SAR.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
Using transfer learning to detect galaxy mergers Ackermann, Sandro; Schawinski, Kevin; Zhang, Ce ...
Monthly Notices of the Royal Astronomical Society,
09/2018, Volume:
479, Issue:
1
Journal Article
Lung cancer is the most prevalent cancer worldwide, and histopathological assessment is indispensable for its diagnosis. However, human evaluation of pathology slides cannot accurately predict ...patients' prognoses. In this study, we obtain 2,186 haematoxylin and eosin stained histopathology whole-slide images of lung adenocarcinoma and squamous cell carcinoma patients from The Cancer Genome Atlas (TCGA), and 294 additional images from Stanford Tissue Microarray (TMA) Database. We extract 9,879 quantitative image features and use regularized machine-learning methods to select the top features and to distinguish shorter-term survivors from longer-term survivors with stage I adenocarcinoma (P<0.003) or squamous cell carcinoma (P=0.023) in the TCGA data set. We validate the survival prediction framework with the TMA cohort (P<0.036 for both tumour types). Our results suggest that automatically derived image features can predict the prognosis of lung cancer patients and thereby contribute to precision oncology. Our methods are extensible to histopathology images of other organs.
The technology of "through-the-wall radar imaging" enables seeing through walls by reconstructing the image based on backscattered waves passing through an optically opaque dielectric wall. However, ...when the electromagnetic waves cannot penetrate the shielding walls (such as metallic walls), a problem called "hard-wall radar imaging" (HWRI) must be investigated. As the objects are located in the shadow region, where only the diffracted field can be reflected back via edges, conventional imaging techniques are not applicable. The HWRI problem is an imaging scenario with an intrinsic multipath of signal propagation. Therefore, it has been studied using a model-based time-reversal (TR) decomposition of TR operator imaging method. In this paper, an imaging technique with a distributed transmitting and receiving array is proposed as a robust way resolving the problem with a low signal-to-clutter ratio in TR imaging. In this method, the interelement response matrix is collected for the construction of the covariance matrix, from which the geometry of the hard wall is determined from the direction of arrival by the multiple signal classification techniques. Next, the adaptive interference cancellation technique is applied to suppress the interference terms in the image function due to the multipath effects. The experimental results were obtained from a distributed multiple-input multiple-output radar system and show that this method is useful not only for the detection of a weak object echo but also for the localization of multiple targets hidden behind the wall.