Density Information Research Articles

Efficient DNA-based data storage using shortmer combinatorial encoding

Data storage in DNA has recently emerged as a promising archival solution, offering space-efficient and long-lasting digital storage solutions. Recent studies suggest leveraging the inherent redundancy of synthesis and sequencing technologies by using composite DNA alphabets. A major challenge of this approach involves the noisy inference process, obstructing large composite alphabets. This paper introduces a novel approach for DNA-based data storage, offering, in some implementations, a 6.5-fold increase in logical density over standard DNA-based storage systems, with near-zero reconstruction error. Combinatorial DNA encoding uses a set of clearly distinguishable DNA shortmers to construct large combinatorial alphabets, where each letter consists of a subset of shortmers. We formally define various combinatorial encoding schemes and investigate their theoretical properties. These include information density and reconstruction probabilities, as well as required synthesis and sequencing multiplicities. We then propose an end-to-end design for a combinatorial DNA-based data storage system, including encoding schemes, two-dimensional (2D) error correction codes, and reconstruction algorithms, under different error regimes. We performed simulations and show, for example, that the use of 2D Reed-Solomon error correction has significantly improved reconstruction rates. We validated our approach by constructing two combinatorial sequences using Gibson assembly, imitating a 4-cycle combinatorial synthesis process. We confirmed the successful reconstruction, and established the robustness of our approach for different error types. Subsampling experiments supported the important role of sampling rate and its effect on the overall performance. Our work demonstrates the potential of combinatorial shortmer encoding for DNA-based data storage and describes some theoretical research questions and technical challenges. Combining combinatorial principles with error-correcting strategies, and investing in the development of DNA synthesis technologies that efficiently support combinatorial synthesis, can pave the way to efficient, error-resilient DNA-based storage solutions.

A new surrogate method for the neutron kinetics calculation of nuclear reactor core transients

Reactor core transient calculation is very important for the reactor safety analysis, in which the kernel is neutron kinetics calculation by simulating the variation of neutron density or thermal power over time. Compared with the point kinetics method, the time-space neutron kinetics calculation can provide accurate variation of neutron density in both space and time domain. But it consumes a lot of resources. It is necessary to develop a surrogate model that can quickly obtain the temporal and spatial variation information of neutron density or power with acceptable calculation accuracy. This paper uses the time-varying characteristics of power to construct a time function, parameterizes the time-varying characteristics which contains the information about the spatial change of power. Thereby, the amount of targets to predict in the space domain is compressed. A surrogate method using the machine learning is proposed in this paper. In the construction of a neural network, the input is processed by a convolutional layer, followed by a fully connected layer or a deconvolution layer. For the problem of time sequence disturbance, a structure combining convolutional neural network and recurrent neural network is used. It is verified in the tests of a series of 1D, 2D and 3D reactor models. The predicted values obtained using the constructed neural network models in these tests are in good agreement with the reference values, showing the powerful potential of the surrogate models.

Wasserstein distance-based full waveform inversion method for density reconstruction

Reliable density information is an essential factor in lithologic interpretation and reservoir evaluation. However, density reconstruction is hampered by a lack of long-wavelength information and the crosstalk between different parameters. The powerful nonlinearity in density inversion constrains the multi-parameter full waveform inversion application. This paper introduces the Wasserstein distance into the full waveform inversion for velocity and density based on the optimal transport theory. We construct an objective function with better convexity to avoid the cycle-skipping caused by the poor initial model. Furthermore, due to the sensitivity of the quadratic Wasserstein distance to low-frequency, we add the low-wavenumber component absent in the conventional density inversion. We acquire a long-wavelength density background using optimal transport inversion and then apply it as an initial model for least-squares inversion. Numerical examples show that our optimal transport inversion effectively builds a reasonable density background model for FWI. Density inversion can be improved by incorporating optimal transport and least-squares theory, leading to a reliable quantitative description for petroleum exploration and development.

The ℓ1−2-norm-regularized basis pursuit seismic inversion based on the exact Zoeppritz equation

Sparsity constraints have been widely adopted in the regularization of ill-posed problems to obtain subsurface properties with sparseness features. However, the target parameters are generally not sparsely distributed, and sparsity constraints lead to results that are missing information. Smooth constraints (e.g., [Formula: see text] norm) lead to the insufficient resolution of the inversion results. To overcome this issue, an effective solution is to convert the target parameters to a sparse representation, which can then be solved with sparsity constraints. For the estimation of elastic parameters, a high-resolution and reliable seismic basis pursuit inversion is developed based on the exact Zoeppritz equation. Furthermore, the [Formula: see text] norm is developed as a constraint, wherein a regularized function is minimized with the alternating direction method of the multipliers algorithm. Numerical examples and real data applications demonstrate that our method can not only improve the accuracy of the inversion results, especially the S-wave velocity and density information but also increase the resolution of the inversion results. Furthermore, the [Formula: see text]-norm constraint has better noise suppression demonstrating great potential in practical applications.

What Does That Mean? Complementizers and Epistemic Authority.

A core goal of research in language is to understand the factors that guide choice of linguistic form where more than one option is syntactically well-formed. We discuss one case of optionality that has generated longstanding discussion: the choice of either using or dropping the English complementizer that in sentences like I think (that) the cat followed the dog. Existing psycholinguistic analyses tie that-usage to production pressures associated with sentence planning (Ferreira & Dell, 2000), avoidance of ambiguity (Hawkins, 2004), and relative information density (Jaeger, 2010). Building on observations from cross-linguistic fieldwork, we present a novel proposal in which English that can serve to mark a speaker's "epistemic authority" over the information packaged within the embedded clause; that is, it indicates that the speaker has more knowledge of the embedded proposition compared with their addressee and thus has a perspective that they believe their addressee doesn't share. Testing this proposal with a forced-choice task and a series of corpus surveys, we find that English that is keyed to the use of embedded speaker (first-person) subject pronouns and occurs in sentences containing newsworthy information. Our account of that-optionality takes into account why that is associated with both (i) a dense information signal and (ii) semantic-pragmatic content, as well as extending to cases of non-optionality in subject/sentence-initial clauses (e.g., *(That) the cat is following the dog, I already know) and fragment answers (e.g., What do you already know? *(That) the cat is following the dog), where that is required.

DNA Information Storage and Cryptography System

With the development of information technology, the global data volume is growing exponentially. In order to alleviate the contradiction between massive data and traditional storage technology, people begin to seek for a new generation of storage media. As a carrier of genetic information, DNA has the characteristics of high information density, long storage life and low maintenance cost, which can effectively overcome the deficiency of traditional storage media. With the development of DNA synthesis and DNA sequencing technology, DNA data storage technology has attracted more and more attention, and a series of major breakthroughs have been made. In this paper, with the workflow of DNA data storage as the main line, expounds the basic theory of DNA data storage and related technology, mainly introduced the research progress of DNA storage method and strategy, briefly summarizes the latest research results of DNA data cryptography, and finally discussed the major challenges that DNA data storage technology is facing , especially, DNA synthesis efficiency, DNA sequencing time cost and DNA data cryptography will be an important research direction of DNA storage technology in the future. It is believed that with the deepening of the data storage research on DNA, DNA data storage will become the most potential new storage method in the future, and can be a practical application in the future.

Statistical analysis of magnetic divertor configuration influence on H-mode transitions

DIII-D plasmas are compared for two upper divertor configurations: with the outer strike point on the small angle slot (SAS) divertor target and with the outer strike point on the horizontal divertor target (HT). Scanning the vertical distance between the magnetic null point and the divertor target over a range 0.10–0.16 m is shown to increase the threshold power, Pth , and edge plasma power, PLoss , for the low-to-high confinement (L–H) and H–L transitions respectively, by up to a factor of 1.4. The X-point height scans were performed at three L-mode core plasma line average electron densities, n¯e= 1.2, 2.2 and 3.6 ×1019m−3 , to investigate the density dependence of divertor magnetic configuration influence on Pth . The X-point height, Zx-pt , was further extended across the range 0.16–0.22 m with the more open HT divertor configuration, for which a clear decrease in Pth with increasing Zx-pt is observed. The dependence of Pth on divertor magnetic geometry is further investigated using a time-dependent probability density function (PDF) model and information geometry to elucidate the roles played by pedestal plasma turbulence and perpendicular velocity flows. The degree of stochasticity of the plasma turbulence is observed to be sensitive to the plasma heating rate. The calculated square of the information rate shows changes in the relative density fluctuations and perpendicular velocity PDFs begin 2–5 ms prior to the L–H transition for three plasmas; providing a crucial measurement of the dynamic timescale of external transport barrier formation. Additionally, both information length and rate provide potential predictors of the L–H transition for these plasmas.

Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training

The goal of this paper is to alleviate the training cost for few-shot semantic segmentation (FSS) models. Despite that FSS in nature improves model generalization to new concepts using only a handful of test exemplars, it relies on strong supervision from a considerable amount of labeled training data for base classes. However, collecting pixel-level annotations is notoriously expensive and time-consuming, and small-scale training datasets convey low information density that limits test-time generalization. To resolve the issue, we take a pioneering step towards label-efficient training of FSS models from fully unlabeled training data, or additionally a few labeled samples to enhance the performance. This motivates an approach based on a novel unsupervised meta-training paradigm. In particular, the approach first distills pre-trained unsupervised pixel embedding into compact semantic clusters from which a massive number of pseudo meta-tasks is constructed. To mitigate the noise in the pseudo meta-tasks, we further advocate a robust Transformer-based FSS model with a novel prototype-based cross-attention design. Extensive experiments have been conducted on two standard benchmarks, i.e., PASCAL-5i and COCO-20i, and the results show that our method produces impressive performance without any annotations, and is comparable to fully supervised competitors even using only 20% of the annotations. Our code is available at: https://github.com/SSSKYue/UMTFSS.

The Prognostic Quality of Risk Prediction Models to Assess the Individual Breast Cancer Risk in Women: An Overview of Reviews

Purpose. Breast cancer is the most common cancer among women globally, with an incidence of approximately two million cases in 2018. Organised age-based breast cancer screening programs were established worldwide to detect breast cancer earlier and to reduce mortality. Currently, there is substantial anticipation regarding risk-adjusted screening programs, considering various risk factors in addition to age. The present study investigated the discriminatory accuracy of breast cancer risk prediction models and whether they suit risk-based screening programs. Methods. Following the PICO scheme, we conducted an overview of reviews and systematically searched four databases. All methodological steps, including the literature selection, data extraction and synthesis, and the quality appraisal were conducted following the 4-eyes principle. For the quality assessment, the AMSTAR 2 tool was used. Results. We included eight systematic reviews out of 833 hits based on the prespecified inclusion criteria. The eight systematic reviews comprised ninety-nine primary studies that were also considered for the data analysis. Three systematic reviews were assessed as having a high risk of bias, while the others were rated with a moderate or low risk of bias. Most identified breast cancer risk prediction models showed a low prognostic quality. Adding breast density and genetic information as risk factors only moderately improved the models’ discriminatory accuracy. Conclusion. All breast cancer risk prediction models published to date show a limited ability to predict the individual breast cancer risk in women. Hence, it is too early to implement them in national breast cancer screening programs. Relevant randomised controlled trials about the benefit-harm ratio of risk-adjusted breast cancer screening programs compared to conventional age-based programs need to be awaited.

Design of unbalanced 9:2 ternary encoder and 2:9 ternary decoder circuits in resistive random access memory and carbon nanotube field effect transistor technology

SummaryMultivalued logic (MVL) offers higher information density as compared with binary logic systems for an equal number of digits. In a ternary microprocessor, memory access is the most time‐ and power‐consuming action. In order to design a ternary computer, the possibilities for designing ternary encoders and decoders must be examined. This paper presents the designs of 9:2 unbalanced ternary encoder (TENC) and 2:9 unbalanced ternary line address decoder (TDEC) based on novel designs of standard ternary inverter (STI), ternary NAND (TNAND), and ternary NOR (TNOR) logic gates using resistive random‐access memory (RRAM) and carbon nanotube field effect transistor (CNTFET) technology. The simulation results indicate an improvement in performance parameters such as power consumption, delay, power–delay product (PDP), and component count of the proposed circuits in comparison with the different existing counterparts. Moreover, a detailed analysis is carried out on the impact of different process, voltage, and temperature (PVT) variations on the performance metrics, including propagation delay, power consumption, and PDP of the 9:2 unbalanced ternary encoder and 2:9 unbalanced ternary line address decoder. The proposed ternary encoder circuit shows an improvement of 62% in delay, 71% in power consumption, 88.6% improvement in PDP, and 53% reduction in CNTFET count, whereas the ternary decoder circuit exhibits an improvement of 6.6% and 94% in delay and 43% and 94% improvement in power consumption, respectively. Moreover, the CNTFET count is reduced by 16.07% and 46%.

Growth of multiferroic γ-BaFe2O4 thin films by Pulsed Electron Deposition technique

Multiferroic materials have attracted significant attention due to their potential to revolutionize logic and memory devices by reducing energy consumption or increasing information density. Barium monoferrite (γ-BaFe2O4 or BaFeO) is a promising lead- and bismuth-free multiferroic material with a stuffed tridymite-like structure, exhibiting both ferroelectricity and antiferromagnetism at room temperature. BaFeO thin films with engineered grain orientation, composition, elastic strain, and magnetic interlayer couplings could open new avenues for novel devices. However, synthesizing the stoichiometric "1–2–4" phase in the BaO-Fe2O3 system remains a challenge. This study focuses on the deposition of pure and crystalline γ-BaFe2O4 thin films using the Pulsed Electron Deposition (PED) method. An analysis of the plasma plume revealed that two main mechanisms govern the dynamics of BaFeO deposition: a congruent ablation process occurring far from thermodynamic equilibrium and an incongruent low-energy evaporation process. The regions of the film with maximized ablative/evaporative ratios are free from the secondary Ba2Fe2O5 phase, suggesting that the thermal evaporative contribution is the primary factor responsible for the out-of-stoichiometry composition. The influence of the substrate type on the structural properties of BaFeO films was also evaluated. It was confirmed that (h00)-oriented BaFeO tends to grow on amorphous quartz and polycrystalline Pt-coated quartz in the 700–800°C temperature range. Conversely, the growth on SiO2/Si leads to the preferential grain orientation along the (0k0) and (0kl) directions by increasing the temperature. These findings demonstrate the potential of PED for producing high-quality γ-BaFe2O4 polycrystalline thin films. A fully heteroepitaxial (h00)-BaFeO/(111)-MgO system was even achieved by depositing on a single-crystal MgO substrate. This unprecedented result paves the way for utilizing such thin films in multiferroic applications, where phase purity, structural orientation, and defect-free structures can be exploited to enhance the magnetoelectric properties of the device.

Coronin 1-dependent cell density sensing and regulation of the peripheral T cell population size.

The establishment and maintenance of peripheral T cells is important to ensure appropriate immunity. In mammals, T cells are produced in the thymus before seeding the periphery early in life, and thereafter progressive thymus involution impairs new T cell production. Yet, peripheral T cells are maintained lifelong at approximately similar cell numbers. The question thus arises: what are the mechanisms that enable the maintenance of the appropriate number of circulating T cells, ensuring that T cell numbers are neither too low nor too high? Here, we highlight recent research suggesting a key role for coronin 1, a member of the evolutionarily conserved family of coronin proteins, in both allowing T cells to reach as well as maintain their appropriate cell population size. This cell population size controlling pathway was found to be conserved in amoeba, mice and human. We propose that coronin 1 is an integral part of a cell-intrinsic pathway that couples cell density information with prosurvival signalling thereby regulating the appropriate number of peripheral T cells.

Channel-wise attention enhanced and structural similarity constrained cycleGAN for effective synthetic CT generation from head and neck MRI images

BackgroundMagnetic resonance imaging (MRI) plays an increasingly important role in radiotherapy, enhancing the accuracy of target and organs at risk delineation, but the absence of electron density information limits its further clinical application. Therefore, the aim of this study is to develop and evaluate a novel unsupervised network (cycleSimulationGAN) for unpaired MR-to-CT synthesis.MethodsThe proposed cycleSimulationGAN in this work integrates contour consistency loss function and channel-wise attention mechanism to synthesize high-quality CT-like images. Specially, the proposed cycleSimulationGAN constrains the structural similarity between the synthetic and input images for better structural retention characteristics. Additionally, we propose to equip a novel channel-wise attention mechanism based on the traditional generator of GAN to enhance the feature representation capability of deep network and extract more effective features. The mean absolute error (MAE) of Hounsfield Units (HU), peak signal-to-noise ratio (PSNR), root-mean-square error (RMSE) and structural similarity index (SSIM) were calculated between synthetic CT (sCT) and ground truth (GT) CT images to quantify the overall sCT performance.ResultsOne hundred and sixty nasopharyngeal carcinoma (NPC) patients who underwent volumetric-modulated arc radiotherapy (VMAT) were enrolled in this study. The generated sCT of our method were more consistent with the GT compared with other methods in terms of visual inspection. The average MAE, RMSE, PSNR, and SSIM calculated over twenty patients were 61.88 ± 1.42, 116.85 ± 3.42, 36.23 ± 0.52 and 0.985 ± 0.002 for the proposed method. The four image quality assessment metrics were significantly improved by our approach compared to conventional cycleGAN, the proposed cycleSimulationGAN produces significantly better synthetic results except for SSIM in bone.ConclusionsWe developed a novel cycleSimulationGAN model that can effectively create sCT images, making them comparable to GT images, which could potentially benefit the MRI-based treatment planning.

DNA-QLC: an efficient and reliable image encoding scheme for DNA storage

BackgroundDNA storage has the advantages of large capacity, long-term stability, and low power consumption relative to other storage mediums, making it a promising new storage medium for multimedia information such as images. However, DNA storage has a low coding density and weak error correction ability.ResultsTo achieve more efficient DNA storage image reconstruction, we propose DNA-QLC (QRes-VAE and Levenshtein code (LC)), which uses the quantized ResNet VAE (QRes-VAE) model and LC for image compression and DNA sequence error correction, thus improving both the coding density and error correction ability. Experimental results show that the DNA-QLC encoding method can not only obtain DNA sequences that meet the combinatorial constraints, but also have a net information density that is 2.4 times higher than DNA Fountain. Furthermore, at a higher error rate (2%), DNA-QLC achieved image reconstruction with an SSIM value of 0.917.ConclusionsThe results indicate that the DNA-QLC encoding scheme guarantees the efficiency and reliability of the DNA storage system and improves the application potential of DNA storage for multimedia information such as images.

Density-Based Discriminative Nonnegative Representation Model for Imbalanced Classification

Representation-based methods have found widespread applications in various classification tasks. However, these methods cannot deal effectively with imbalanced data scenarios. They tend to neglect the importance of minority samples, resulting in bias toward the majority class. To address this limitation, we propose a density-based discriminative nonnegative representation approach for imbalanced classification tasks. First, a new class-specific regularization term is incorporated into the framework of a nonnegative representation based classifier (NRC) to reduce the correlation between classes and improve the discrimination ability of the NRC. Second, a weight matrix is generated based on the hybrid density information of each sample’s neighbors and the decision boundary, which can assign larger weights to minority samples and thus reduce the preference for the majority class. Furthermore, the resulting model can be efficiently optimized through the alternating direction method of multipliers. Extensive experimental results demonstrate that our proposed method is superior to numerous state-of-the-art imbalanced learning methods.

The impact of the digital economy on high-quality agricultural development--Based on the regulatory effects of financial development.

The digital economy, as a new economic form with high information density, provides a new driving force for the realization of high-quality agricultural development. Panel data of 31 provinces in China from 2011 to 2020 were selected for analysis. The static panel data interaction effect model and panel threshold model were used to verify the nonlinear influence mechanism and heterogeneity of financial development in the process of the digital economy affecting high-quality agricultural development. The findings are as follows. (1) During the study period, the high-quality development of China's agriculture showed a steady upward trend; however, the regional differences were significant, and the eastern part was larger than the central and western part. (2) The digital economy can promote high-quality agricultural development. (3) The digital economy has a double threshold effect in the process of affecting high-quality agricultural development, which depends on the level of financial development. When the threshold is exceeded, the digital economy has a more significant promoting effect on high-quality agricultural development. (4) The impact of the digital economy on high-quality agricultural development is heterogeneous. From the perspective of different regions, the impact effect is greatest in the eastern region, while the effect is smaller in the central and western regions. From different resource endowments, the positive impact effect is greatest in the major grain-selling areas, followed by the major grain producing areas, but the positive digital economy driving effect is not significant in the balance of production and sales areas. Finally, three policy suggestions are proposed. First, the Chinese government should increase investments in and support for digital technology to promote the integration of the digital economy and agriculture. Second, the Chinese government should promote the development of digital inclusive finance in areas with financial development below the threshold. Third, different regions should implement differentiated digital economies to promote high-quality agricultural development.

Signal Smoothing and Syntactic Choices: A Critical Reflection on the UID Hypothesis.

The Smooth Signal Redundancy Hypothesis explains variations in syllable length as a means to more uniformly distribute information throughout the speech signal. The Uniform Information Density hypothesis seeks to generalize this to choices on all linguistic levels, particularly syntactic choices. While there is some evidence for the Uniform Information Density hypothesis, it faces several challenges, four of which are discussed in this paper. First, it is not clear what exactly counts as uniform. Second, there are syntactic alternations that occur systematically but that can cause notable fluctuations in the information signature. Third, there is an increasing body of negative results. Fourth, there is a lack of large-scale evidence. As to the fourth point, this paper provides a broader array of data-936 sentence pairs for nine syntactic constructions-and analyzes them in a test setup that treats the hypothesis as a classifier. For our data, the Uniform Information Density hypothesis showed little predictive capacity. We explore ways to reconcile our data with theory.

Enhancing pseudo label quality for pedestrian and cyclist in weakly supervised 3D object detection

Weakly supervised 3D object detection for autonomous driving primarily focuses on cars because of their distinct rectangle boundaries and abundant instances. However, detecting categories with ambiguous rectangle boundaries and fewer instances than cars, such as pedestrians and cyclists, remains challenging with limited research. Ambiguity in rectangle boundaries presents significant difficulties in generating accurate 3D pseudo labels, while the scarcity of instances often leads to convergence issues during detector training. Pedestrians and cyclists are dense inside the 3D bounding boxes but sparse at corners and boundaries. Density is a practical clue for locating and discriminating pedestrians and cyclists in point clouds. This paper proposes a density-based 3D pseudo-label generation module (DPL-3D), addressing the challenges of ambiguous rectangle boundaries. Ambiguity rectangle boundaries will lead to poor pseudo-label quality. Therefore, By leveraging the density information of 3D points, our DPL-3D improves the accuracy and localization quality of the generated pseudo labels. It effectively segments background points, improving the estimation of pseudo labels’ location, dimension, and orientation. Few training samples always lead to local optima. Introducing multi-modal data in the detector network could enhance the constraints of objects’ features, but 2D images and 3D point clouds have a resolution gap. A motivation for dealing with the resolution gap is that neighboring regions with similar colors and textures in 2D images may exhibit spatial proximity in 3D space. Therefore, a multi-modal network driven by superpixel segmentation is introduced. This network enables effective discrimination between objects in 2D images and 3D point clouds, bridging the resolution gap and leveraging complementary features from both modalities. Experimental results on the KITTI dataset demonstrate the effectiveness of the proposed methods in addressing the challenges associated with weakly-supervised 3D object detection, particularly for categories with ambiguous rectangle boundaries and few instances.

Development of a double-grating differential interferometer for plasma diagnostics.

A special differential interferometer consisting of two gratings was developed for diagnostics of plasma density. Compared with other differential interferometers, our system has an important advantage that the shear distance, shear direction, and fringe width can be adjusted independently, enabling easy control of the parameters. This feature allows precise tuning of the two probe beams in the interferometer for rigorous differential phase diagnosis and more accurate information of the plasma density can be obtained. The double-grating-based differential interferometer was tested for diagnostics of the laser-produced plasma which was generated by focusing a 1 TW/35 fs Ti:sapphire laser pulse in a gas jet with a 100 μm orifice diameter. It was confirmed that our differential interferometer can provide more reliable and accurate plasma density information, especially for plasmas with a high spatial gradient in density.

Toward Pixel-Level Precision for Binary Super-Resolution With Mixed Binary Representation.

Binary neural network (BNN) is an effective method for reducing model computational and memory cost, which has achieved much progress in the super-resolution (SR) field. However, there is still a noticeable performance gap between a binary SR network and its full-precision counterpart. Considering that the information density in quantization features is far lower than full-precision features, we aim to improve the precision of quantization features to produce rich-enough output activations for SR task. First, we make several observations that a multibit value could be approximated by multiple 1-bit values, and the computation power of binary convolution could be improved by approximating the multibit convolution process. Then, we propose a mixed binary representation set to approximate multibit activations, which is effective in compensating the quantization precision loss. Finally, we present a new precision-driven binary convolution (PDBC) module, which increases the convolution precision and protects image detail information without extra computation. Compared with normal binary convolution, our method could largely reduce the information loss caused by binarization. In experiments, our methods consistently show superior performance over the baseline models and can surpass state-of-the-art methods in terms of peak signal to noise ratio (PSNR) and visual quality.