Low-resolution Sampling Research Articles

Demonstration and optimization of coherent Doppler wind LiDAR with low sampling resolution

A low sampling resolution scheme for coherent Doppler wind LiDAR (CDWL) is proposed. The CDWL offers advantages in precision and detection resolution but suffers from the requirement of high-speed data acquisition (DAQ) with high sampling resolution, such as 12- or 14-bit, which leads to an increase of the computational complexity and the system cost. The use of a DAQ system with lower sampling resolution can provide a solution to mitigate this problem. The feasibility of the proposed scheme is validated by simulations and experiments. The detection performance can be greatly affected by the quantization interval selected during sampling. It is shown that the optimal quantization interval exists and only depends on the carrier-to-noise ratio (CNR), and the optimal quantization intervals of several sampling resolutions are given at different CNRs. With the given optimal quantization configuration, the low sampling resolution data can be used for reliable wind field measurements. For long-distance detection with a CNR lower than −13dB, the CNR deterioration of 1-bit, 2-bit, 3-bit, and 4-bit signals can be as low as 2, 0.5, 0.2, and 0.1 dB.

A novel signal resolution enhance method based on CBAM-ResNet for bearing intelligent fault diagnosis

Abstract In health condition monitoring of mechanical equipment, the signal is the key source of information. However, signal resolution is often degraded due to factors such as equipment vibration and electromagnetic interference. To address this issue, an Efficient Sub-pixel Convolutional Attention Residual Network (ESPCARN) built on the idea of signal resolution improvement is proposed in this paper. Firstly, the original low-resolution samples are input into a CBAM-ResNet to obtain more feature information of the channels and space within the residual connection and a multi-feature mapping with four channels was generated. Subsequently, the four-channel low-resolution features are aligned periodically through sub-pixel convolution layer, resulting in a set of high-resolution samples and the feature dimension of the data was increased to four times that of the original low-resolution data, thereby realizing the resolution enhancement. Finally, two experiments with different working conditions are established to evaluate the performance of the proposed fault diagnosis method, and the experimental results verified the efficacy of the ESPCARN framework.

Bayesian modelling of time series data (BayModTS)-a FAIR workflow to process sparse and highly variable data.

Systems biology aims to better understand living systems through mathematical modelling of experimental and clinical data. A pervasive challenge in quantitative dynamical modelling is the integration of time series measurements, which often have high variability and low sampling resolution. Approaches are required to utilize such information while consistently handling uncertainties. We present BayModTS (Bayesian modelling of time series data), a new FAIR (findable, accessible, interoperable, and reusable) workflow for processing and analysing sparse and highly variable time series data. BayModTS consistently transfers uncertainties from data to model predictions, including process knowledge via parameterized models. Further, credible differences in the dynamics of different conditions can be identified by filtering noise. To demonstrate the power and versatility of BayModTS, we applied it to three hepatic datasets gathered from three different species and with different measurement techniques: (i) blood perfusion measurements by magnetic resonance imaging in rat livers after portal vein ligation, (ii) pharmacokinetic time series of different drugs in normal and steatotic mice, and (iii) CT-based volumetric assessment of human liver remnants after clinical liver resection. The BayModTS codebase is available on GitHub at https://github.com/Systems-Theory-in-Systems-Biology/BayModTS. The repository contains a Python script for the executable BayModTS workflow and a widely applicable SBML (systems biology markup language) model for retarded transient functions. In addition, all examples from the paper are included in the repository. Data and code of the application examples are stored on DaRUS: https://doi.org/10.18419/darus-3876. The raw MRI ROI voxel data were uploaded to DaRUS: https://doi.org/10.18419/darus-3878. The steatosis metabolite data are published on FairdomHub: 10.15490/fairdomhub.1.study.1070.1.

Degradation regression with uncertainty for blind super-resolution

Some recent blind super-resolution (SR) efforts focus on designing complex degradation models to better simulate real-world degradations. The paired high-resolution (HR) & low-resolution (LR) samples synthesized by these models can cover a large degradation space, which helps train a robust SR model in real scenarios. However, these diverse synthetic samples may render the SR model degradation-unaware and prevent it from achieving optimal results on LR images with specific degradations. Alternatively, another category of methods is proposed to estimate specific degradations in the given application and then tailor a degradation-aware SR model accordingly. Nonetheless, degradation estimation is an ill-posed problem and accurate estimation is quite challenging. Towards these issues, we propose a probabilistic degradation estimator (PDE) which can predict the degradation as a certain distribution rather than a single point. Specifically, we develop an intersection over union (IoU) based degradation regression loss with uncertainty, which could lead PDE to shrink the possible degradation space of the test LR image. This enables the degradation model to synthesize more degradation-specific training samples and further improve SR performance. In this way, our PDE can alleviate degradation redundancy compared with degradation-unaware methods and is more robust to the degradation estimation error than previous degradation-aware methods. Extensive experiments show that the proposed PDE can help the SR model produce better results on both synthetic and real-world images.

Spatial variation and controls of soil microbial necromass carbon in a tropical montane rainforest

Soil microbial necromass carbon is an important component of the soil organic carbon (SOC) pool which helps to improve soil fertility and texture. However, the spatial pattern and variation mechanisms of fungal- and bacterial-derived necromass carbon at local scales in tropical rainforests are uncertain. This study showed that microbial necromass carbon and its proportion in SOC in tropical montane rainforest exhibited large spatial variation and significant autocorrelation, with significant high-high and low-low clustering patterns. Microbial necromass carbon accounted for approximately one-third of SOC, and the fungal-derived microbial necromass carbon and its proportion in SOC were, on average, approximately five times greater than those of bacterial-derived necromass. Structural equation models indicated that soil properties (SOC, total nitrogen, total phosphorus) and topographic features (elevation, convexity, and aspect) had significant positive effects on microbial necromass carbon concentrations, but negative effects on its proportions in SOC (especially the carbon:nitrogen ratio). Plant biomass also had significant negative effects on the proportion of microbial necromass carbon in SOC, but was not correlated with its concentration. The different spatial variation mechanisms of microbial necromass carbon and their proportions in SOC are possibly related to a slower accumulation rate of microbial necromass carbon than of plant-derived organic carbon. Geographic spatial correlations can significantly improve the microbial necromass carbon model fit, and low sampling resolution may lead to large uncertainties in estimating soil carbon dynamics at specific sites. Our work will be valuable for understanding microbial necromass carbon variation in tropical forests and soil carbon prediction model construction with microbial participation.

3D-2D Distance Maps Conversion Enhances Classification of Craniosynostosis.

Diagnosis of craniosynostosis using photogrammetric 3D surface scans is a promising radiation-free alternative to traditional computed tomography. We propose a 3D surface scan to 2D distance map conversion enabling the usage of the first convolutional neural networks (CNNs)-based classification of craniosynostosis. Benefits of using 2D images include preserving patient anonymity, enabling data augmentation during training, and a strong under-sampling of the 3D surface with good classification performance. The proposed distance maps sample 2D images from 3D surface scans using a coordinate transformation, ray casting, and distance extraction. We introduce a CNNbased classification pipeline and compare our classifier to alternative approaches on a dataset of 496 patients. We investigate into low-resolution sampling, data augmentation, and attribution mapping. Resnet18 outperformed alternative classifiers on our dataset with an F1-score of 0.964 and an accuracy of 98.4 %. Data augmentation on 2D distance maps increased performance for all classifiers. Under-sampling allowed 256-fold computation reduction during ray casting while retaining an F1-score of 0.92. Attribution maps showed high amplitudes on the frontal head. We demonstrated a versatile mapping approach to extract a 2D distance map from the 3D head geometry increasing classification performance, enabling data augmentation during training on 2D distance maps, and the usage of CNNs. We found that low-resolution images were sufficient for a good classification performance. Photogrammetric surface scans are a suitable craniosynostosis diagnosis tool for clinical practice. Domain transfer to computed tomography seems likely and can further contribute to reducing ionizing radiation exposure for infants.

An extended Teager study of vibration detection for industrial robot under low sampling resolution

Vibration is an essential indicator that reflects the robust status of industrial robots. To improve the detection effect and resilience for articulated industrial robot in-situ, a novel non-invasive self-diagnostic method is proposed. Feasible embedded motor current is used to calculate the extended Teager indicator at a low sampling rate. Compared with common Teager and other indicators, both theoretical analysis and experimental data indicate the effectiveness and robustness of the proposed method.

Response of Dissolved Trace Metals to Dust Storms, Sediment Resuspension, and Flash Floods in Oligotrophic Oceans

AbstractTrace metals (TM) delivered by atmospheric dust play a key role in oceanic biogeochemical cycles. However, the impact of short‐term environmental perturbations such as dust storms and sediment resuspension events on the oceanic water column is poorly constrained due to the low temporal sampling resolution and episodic nature of these events. The Gulf of Aqaba (GoA), Red Sea, is a highly accessible deep oligotrophic water body featuring exceptionally high atmospheric deposition fluxes that provide the main source of TMs to the GoA surface water. Here, we present a 2‐year time series of dissolved manganese, cobalt, nickel, copper, zinc, cadmium, and phosphate concentration profiles sampled in the GoA. The study focuses on daily time scale dust storms and episodes of sediment resuspension to quantify the immediate impact of these events on dissolved TM cycling. Counter‐intuitively, upper mixed layer TM inventories decrease with increasing aerosol loads, with the effects of aerosol‐induced TM scavenging and dissolution peaking 5–6 days after aerosol deposition. Dust storms promote intense TM scavenging, with TM inventories decreasing by up to 44%, but seldom lead to TM enrichment. Similarly, sediment resuspension and flash flood events triggered significant TM scavenging. These findings highlight the potential dual role of atmospheric deposition in the oceans as a long‐term source of dissolved TMs and a short‐term sink. The in situ observations presented here may be used to understand and quantify the global impact of abrupt environmental events on oceanic chemical compositions.

Real-Time Segmentation of Artificial Targets Using a Dual-Modal Efficient Attention Fusion Network

The fusion of spectral–polarimetric information can improve the autonomous reconnaissance capability of unmanned aerial vehicles (UAVs) in detecting artificial targets. However, the current spectral and polarization imaging systems typically suffer from low image sampling resolution, which can lead to the loss of target information. Most existing segmentation algorithms neglect the similarities and differences between multimodal features, resulting in reduced accuracy and robustness of the algorithms. To address these challenges, a real-time spectral–polarimetric segmentation algorithm for artificial targets based on an efficient attention fusion network, called ESPFNet (efficient spectral–polarimetric fusion network) is proposed. The network employs a coordination attention bimodal fusion (CABF) module and a complex atrous spatial pyramid pooling (CASPP) module to fuse and enhance low-level and high-level features at different scales from the spectral feature images and the polarization encoded images, effectively achieving the segmentation of artificial targets. Additionally, the introduction of the residual dense block (RDB) module refines feature extraction, further enhancing the network’s ability to classify pixels. In order to test the algorithm’s performance, a spectral–polarimetric image dataset of artificial targets, named SPIAO (spectral–polarimetric image of artificial objects) is constructed, which contains various camouflaged nets and camouflaged plates with different properties. The experimental results on the SPIAO dataset demonstrate that the proposed method accurately detects the artificial targets, achieving a mean intersection-over-union (MIoU) of 80.4%, a mean pixel accuracy (MPA) of 88.1%, and a detection rate of 27.5 frames per second, meeting the real-time requirement. The research has the potential to provide a new multimodal detection technique for enabling autonomous reconnaissance by UAVs in complex scenes.

Site-specific length-biomass relationships of arctic arthropod families are critical for accurate ecological inferences.

Arthropods play a crucial role in terrestrial ecosystems, for instance in mediating energy fluxes and in forming the food base for many organisms. To better understand their functional role in such ecosystem processes, monitoring of trends in arthropod biomass is essential. Obtaining direct measurements of the body mass of individual specimens is laborious. Therefore, these data are often indirectly acquired by utilizing allometric length-biomass relationships based on a correlative parameter, such as body length. Previous studies have often used such relationships with a low taxonomic resolution and/or small sample size and/or adopted regressions calibrated in different biomes. Despite the scientific interest in the ecology of arctic arthropods, no site-specific family-level length-biomass relationships have hitherto been published. Here we present 27 family-specific length-biomass relationships from two sites in the High Arctic: Zackenberg in northeast Greenland and Knipovich in north Taimyr, Russia. We show that length-biomass regressions from different sites within the same biome did not affect estimates of phenology but did result in substantially different estimates of arthropod biomass. Estimates of daily biomass at Zackenberg were on average 24% higher when calculated using regressions for Knipovich compared to using regressions for Zackenberg. In addition, calculations of daily arthropod biomass at Zackenberg based on order-level regressions from frequently cited studies in literature revealed overestimations of arthropod biomass ranging from 69.7% to 130% compared to estimates based on regressions for Zackenberg. Our results illustrate that the use of allometric relationships from different sites can significantly alter the biological interpretation of, for instance, the interaction between insectivorous birds and their arthropod prey. We conclude that length-biomass relationships should be locally established rather than being based on global relationships.

Using isotopic tracers to enhance routine watershed monitoring – Insights from an intensively managed agricultural catchment

Experimental (research-based) and non-research-based watershed monitoring programs often differ with respect to sampling frequency, monitored variables, and monitoring objectives. Isotopic variables, which are more commonly incorporated in research-based programs, can provide an indication of water sources and the transit time of water in a catchment. These variables may be a valuable complement to traditional water quality monitoring variables and have the potential to support improved hydrologic process-related insights from long term monitoring programs that typically have low resolution sampling. The purpose of this investigation is to explore the utility of incorporating isotopic variables (specifically δ18O, δ2H, and 222Rn) into routine monthly sampling regimes by comparing insights gained from these variables to monitoring only specific conductivity and chloride. A complete annual cycle of monthly groundwater and surface water monitoring data collected from the Upper Parkhill watershed in southwestern Ontario, Canada was used to characterize baseline watershed conditions, evaluate watershed resilience to climate change, and examine contamination vulnerability. Study results provide an improved understanding of appropriate tracer use in agricultural regions with isotopic variables able to provide important insights into the seasonality of hydrologic phenomena, such as the timing of groundwater recharge. A comparison of monitoring variables to present-day hydro-meteorological conditions suggests the importance of a winter dominated hydrologic regime and the potential influence of changes in precipitation on groundwater-surface water interactions. Estimated transit time dynamics indicate the likelihood for rapid contaminant transport through surface and shallow subsurface flow and highlight the possible effects of agricultural tile drainage. The sampling approach and data analysis methods adopted in this study provide the basis for improving routine watershed monitoring programs in agricultural regions.

Downstream Distribution and Postdepositional Mobilization of Cadmium in Alluvial Soils

The geochemical signature in alluvial soils is a witness of human activities that took place in a river catchment. Sampling of alluvial soils at depth, in combination with information on sedimentological history and age of samples, may even allow to reconstruct the pollution history of the river basin. In the present study, data on alluvial soils contaminated by a major pollution source were analyzed, with special attention for these soils as an archive for information on the pollution history of a river/river catchment, and on the postdepositional downward migration of metal(loid)s in the alluvial soils. Besides the lateral variation of soil properties and metal(loid) concentrations in the alluvial soils, the vertical distribution of metal(loid)s in soil profiles, as well as the evolution of soil composition in relation to the distance from the river, was addressed. The postdepositional mobilization of Cd was evaluated in a fine-scale sampled alluvial soil core, by comparing data from 137Cs dating with data about the Cd emissions through time and by using leaching tests to calculate the downward migration of Cd. A substantial amount of Cd could leach from superficial to deeper soil layers. Therefore, the low-resolution (cm-scale) sampling of the alluvial soil was not reliable to reconstruct the pollution history of the river catchment, because the elevated chloride-concentrations in the river water increased the downward leaching of Cd through the formation of chloro-complexes. Moreover, the variability in flooding and sedimentation regimes along the river resulted in a heterogeneous composition of the alluvial soils, allowing very large differences in metal(loid) concentrations in places only a few meters apart.

Snapshot multi-dimensional computational imaging through a liquid crystal diffuser

Multi-dimensional optical imaging systems that simultaneously gather intensity, depth, polarimetric, and spectral information have numerous applications in medical sciences, robotics, and surveillance. Nevertheless, most current approaches require mechanical moving parts or multiple modulation processes and thus suffer from long acquisition time, high system complexity, or low sampling resolution. Here, a methodology to build snapshot multi-dimensional lensless imaging is proposed by combining planar-optics and computational technology, benefiting from sufficient flexibilities in optical engineering and robust information reconstructions. Specifically, a liquid crystal diffuser based on geometric phase modulation is designed to simultaneously encode the spatial, spectral, and polarization information of an object into a snapshot detected speckle pattern. At the same time, a post-processing algorithm acts as a special decoder to recover the hidden information in the speckle with the independent and unique point spread function related to the position, wavelength, and chirality. With the merits of snapshot acquisition, multi-dimensional perception ability, simple optical configuration, and compact device size, our approach can find broad potential applications in object recognition and classification.

Implementation of Generative Adversarial Networks in Mobile Applications for Image Data Enhancement

This article aims to explore and research GANs as a tool for mobile devices that can generate high-resolution images from low-resolution samples and reduce blurring. In addition, the authors also analyse the specifics of GAN, SRGAN, and ESRGAN loss functions and their features. GANs are widely used for a vast range of applied tasks for image manipulations. They’re able to synthesize, combine, and restore graphical samples of high quality that are almost indistinguishable from real data. The main scope of the research is to study the possibility to use GANs for the said tasks, and their potential implementation in mobile applications.

SnapperGPS: Open Hardware for Energy-Efficient, Low-Cost Wildlife Location Tracking with Snapshot GNSS

Location tracking with global navigation satellite systems (GNSS), such as the GPS, is used in many applications, including the tracking of wild animals for research. Snapshot GNSS is a technique that only requires milliseconds of satellite signals to infer the position of a receiver. This is ideal for low-power applications such as animal tracking. However, there are few existing snapshot systems, none of which is open source. To address this, we developed SnapperGPS, a fully open-source, low-cost, and low-power location tracking system designed for wildlife tracking. SnapperGPS comprises three parts, all of which are open-source: (i) a small, low-cost, and low-power receiver; (ii) a web application to configure the receiver via USB; and (iii) a cloud-based platform for processing recorded data. This paper presents the hardware side of this project. The total component cost of the receiver is under $30, making it feasible for field work with restricted budgets and low recovery rates. The receiver records very short and low-resolution samples resulting in particularly low power consumption, outperforming existing systems. It can run for more than a year on a 40 mAh battery. We evaluated SnapperGPS in controlled static and dynamic tests in a semi-urban environment where it achieved median errors of 12 m. Additionally, SnapperGPS has already been deployed for two wildlife tracking studies on sea turtles and sea birds.

Riverine nitrate source and transformation as affected by land use and land cover

A mixed land use/land cover (LULC) catchment increases the complexity of sources and transformations of nitrate in rivers. Spatial paucity of sampling particularly low-resolution sampling in tributaries can result in a bias for identifying nitrate sources and transformations. In this study, high spatial resolution sampling campaigns covering mainstream and tributaries in combination with hydro-chemical parameters and dual isotopes of nitrate were performed to reveal spatio-temporal variations of nitrate sources and transformations in a river draining a mixed LULC catchment. This study suggested that point sources dominated the nitrate in the summer and winter, while non-point sources dominated the nitrate in the spring and autumn. A positive correlation was observed between proportions from sewage and land use index (LUI). However, negative correlations between soil nitrogen/nitrogen fertilizer and LUI were observed. With an increase of urban areas, the increased contribution from domestic sewage resulted in an increase of NO3− concentrations in rivers. Both urban and agricultural inputs should be considered in nitrate pollution management in a mixed LULC catchment. We concluded that the seasonal variations of nitrate sources were mainly affected by flow velocity conditions and agricultural activities, while spatial variations were mainly affected by LULC. In addition, we found a novel underestimation of dominated sources from Bayesian model because of mixing effect of isotope values from the tributaries to mainstream, however, high spatial resolution sampling can make up for this shortcoming. δ15N and δ18O values of nitrate indicated that nitrate originated from nitrification in soils. The nitrate concentrations and correlation between δ15N and 1/[NO3−] suggested little contribution of nitrate removal by denitrification. Thus, the nitrate reduction in the Yuehe River basin needs to be strengthened. The study provides new implications for estimation of nitrate sources and transformations and basis for nitrate reduction in the river with mixed LULC catchment.

The limits of mtDNA analysis for determining the provenance of invasive species: a midwife toad example

Abstract For most if not all European herpetofauna, range-wide mtDNA phylogeographies have been published. This facilitates establishing the provenance of introduced populations. However, precision is contingent on the spatial genetic structure across the range of the taxon under study and, in particular, from where within that range the introduction was sourced. In the Netherlands, the common midwife toad, Alytes obstetricans, only naturally occurs in the extreme southeast and is on the decline there. Yet, introduced populations thrive elsewhere in the country. We use mtDNA analysis to try to determine the origin of two introduced populations along the Dutch coast, in the city of The Hague and the dune area Meijendel. We compiled a database of hundreds of individuals from throughout the distribution range and added over 130 individuals from both native and introduced populations from the Netherlands, Belgium and Germany. The mtDNA haplotypes found in the introduced populations are associated with postglacial expansion. The main haplotype predominates in the natural range in the Netherlands, but also occurs much more widely across western Europe, north of the Pyrenees. A closely related haplotype, newly identified from The Hague, was not found in the native Netherlands range, suggesting an origin from abroad. The combination of low phylogeographic resolution and low sampling density in the postglacially colonized part of the range hampers our ability to determine the provenance of the introduced A. obstetricans populations.

Generating high-resolution climatological precipitation data using SinGAN

ABSTRACT High-resolution (HR) climate data are indispensable for studying regional climate trends, disaster prediction, and urban development planning in the face of climate change. However, state-of-the-art long-term global climate simulations do not provide appropriate HR climate data. Deep learning models are often used to obtain high-resolution climate data. However, due to the fact that these models require sufficient low-resolution (LR) and HR data pairs for the training process, they cannot be applied to scenario with inadequate training data. In this paper, we explore the applicability of a single image generative adversarial network (SinGAN) in generating HR climate data. SinGAN relies on single LR input data to obtain the corresponding HR data. To improve the performance for extreme-value regions, we propose a SinGAN combined with the weighted patchGAN discriminator (WSinGAN). The proposed WSinGAN outperforms comparable models in generating HR precipitation data, and its results are close to real HR data with sharp gradients and more refined small-scale features. We also test the scalability of the pre-trained WSinGAN for unseen samples and show that although only a single LR sample is used to train WSinGAN, it can still produce reliable HR data for unseen data.

A bilateral branch learning paradigm for short term wind power prediction with data of multiple sampling resolutions

This paper revisits the data-driven short-term wind power prediction (WPP) modeling process with introducing a new data usage paradigm, using SCADA data of both high and low sampling resolutions as model inputs. To organically incorporate data of two sampling resolutions as inputs and tackle the extreme high dimension of features induced by using high sampling resolution data in WPP modeling, we propose a novel bilateral branch learning based WPP modeling framework, which includes two data feature engineering branches and one prediction module. One branch for processing the high sampling resolution data is developed based on a novel deep learning structure stacking multiple convolutional and sparse pooling layers as well as long short term memory networks (CNN-SP-LSTM). Another branch adopts the classical feature selection to process low sampling resolution data. The bilateral branch engineered features are concatenated as one input for the final wind power prediction. To justify the validity of the proposed WPP data usage and modeling framework, extensive computational experiments are conducted based on real wind farm SCADA data. Via analyzing computational results, we show that it is advantageous to incorporate data of the high sampling resolution into WPP tasks. Via a comprehensive comparative analytics, we identify the machine learning techniques most suitable on the feature selection and prediction module development parts of the proposed modeling framework. Via an ablation study, we verify the contribution and advantage of the proposed CNN-SP-LSTM module. Based on results of a comprehensive computational study, we verify that our proposed framework achieves the state-of-the art performance as it beats a large set of classical data-driven and recent deep learning based WPP methods considered in this study.

Toward Real-World Super-Resolution via Adaptive Downsampling Models.

Most image super-resolution (SR) methods are developed on synthetic low-resolution (LR) and high-resolution (HR) image pairs that are constructed by a predetermined operation, e.g., bicubic downsampling. As existing methods typically learn an inverse mapping of the specific function, they produce blurry results when applied to real-world images whose exact formulation is different and unknown. Therefore, several methods attempt to synthesize much more diverse LR samples or learn a realistic downsampling model. However, due to restrictive assumptions on the downsampling process, they are still biased and less generalizable. This study proposes a novel method to simulate an unknown downsampling process without imposing restrictive prior knowledge. We propose a generalizable low-frequency loss (LFL) in the adversarial training framework to imitate the distribution of target LR images without using any paired examples. Furthermore, we design an adaptive data loss (ADL) for the downsampler, which can be adaptively learned and updated from the data during the training loops. Extensive experiments validate that our downsampling model can facilitate existing SR methods to perform more accurate reconstructions on various synthetic and real-world examples than the conventional approaches.