Researchers have recently achieved significant advances in deep learning techniques, which in turn has substantially advanced other research disciplines, such as natural language processing, image processing, speech r...
详细信息
Researchers have recently achieved significant advances in deep learning techniques, which in turn has substantially advanced other research disciplines, such as natural language processing, image processing, speech recognition, and software engineering. Various deep learning techniques have been successfully employed to facilitate software engineering tasks, including code generation, software refactoring, and fault localization. Many studies have also been presented in top conferences and journals, demonstrating the applications of deep learning techniques in resolving various software engineering tasks. However,although several surveys have provided overall pictures of the application of deep learning techniques in software engineering,they focus more on learning techniques, that is, what kind of deep learning techniques are employed and how deep models are trained or fine-tuned for software engineering tasks. We still lack surveys explaining the advances of subareas in software engineering driven by deep learning techniques, as well as challenges and opportunities in each subarea. To this end, in this study, we present the first task-oriented survey on deep learning-based software engineering. It covers twelve major software engineering subareas significantly impacted by deep learning techniques. Such subareas spread out through the whole lifecycle of software development and maintenance, including requirements engineering, software development, testing, maintenance, and developer collaboration. As we believe that deep learning may provide an opportunity to revolutionize the whole discipline of software engineering, providing one survey covering as many subareas as possible in software engineering can help future research push forward the frontier of deep learning-based software engineering more systematically. For each of the selected subareas,we highlight the major advances achieved by applying deep learning techniques with pointers to the available datasets i
In the development of static luminescent materials with remarkable optical-thermal performance and low cost, next-generation high-brightness laser lighting faces a key challenge. Herein, a unique composite architectur...
详细信息
In the development of static luminescent materials with remarkable optical-thermal performance and low cost, next-generation high-brightness laser lighting faces a key challenge. Herein, a unique composite architecture of Y3Al5O_(12):Ce^(3+) (YAG) phosphor-in-glass film coated on different heat-conducting substrates (PiGF@HCSs), i.e., PiGF@sapphire, PiGF@Al_(2)O_(3), PiGF@AlN, and PiGF@BN–AlN composites, was designed and prepared by a simple film printing and low-temperature sintering technology. The heat-conducting substrates significantly affect the luminescence saturation and phosphor conversion of PiGF@HCSs, allowing substrates with higher thermal conductivity (TC) to have a higher laser power density (LPD) and higher reflectivity to enable higher luminous efficacy (LE). As a consequence, PiGF@sapphire realizes a luminous flux (LF) of 2076 lm@12 W/mm^(2), which is higher than those of PiGF@Al_(2)O_(3) (1890 lm@15 W/mm^(2)) and PiGF@AlN (1915 lm@24 W/mm^(2)), whilePiGF@BN–AlN enables a maximum LF of 3058 lm@21 W/mm^(2). Furthermore, the LE of PiGF@BN–AlN reaches 194 lm/W, which is 1.6 times that of PiGF@AlN, while those of PiGF@sapphire and PiGF@Al_(2)O_(3) are 192 and 150 lm/W, respectively. The working temperature of PiGF@AlN is only 93.3℃ under LPD of 9 W/mm^(2), while those of PiGF@sapphire, PiGF@Al_(2)O_(3), and PiGF@BN–AlN increase to 193.8, 133.6, and 117℃, respectively. These findings provide guidance for commercial applications of PiGF@HCS converters in high-brightness laser lighting and displays.
Ensemble object detectors have demonstrated remarkable effectiveness in enhancing prediction accuracy and uncertainty quantification. However, their widespread adoption is hindered by significant computational and sto...
详细信息
In the realm of underwater robotics,optical imaging plays a pivotal role in many scientific *** to the effects of absorption and scattering,images captured in turbid water are severely ***,enhancing the quality of und...
详细信息
In the realm of underwater robotics,optical imaging plays a pivotal role in many scientific *** to the effects of absorption and scattering,images captured in turbid water are severely ***,enhancing the quality of underwater optical images stands paramount in ensuring the continued advancement and efficacy of underwater robots across its multifarious applications.
Large-scale neural networks-based federated learning(FL)has gained public recognition for its effective capabilities in distributed ***,the open system architecture inherent to federated learning systems raises concer...
详细信息
Large-scale neural networks-based federated learning(FL)has gained public recognition for its effective capabilities in distributed ***,the open system architecture inherent to federated learning systems raises concerns regarding their vulnerability to potential *** attacks turn into a major menace to federated learning on account of their concealed property and potent destructive *** altering the local model during routine machine learning training,attackers can easily contaminate the global *** detection and aggregation solutions mitigate certain threats,but they are still insufficient to completely eliminate the influence generated by ***,federated unlearning that can remove unreliable models while maintaining the accuracy of the global model has become a *** some existing federated unlearning approaches are rather difficult to be applied in large neural network models because of their high computational ***,we propose SlideFU,an efficient anti-poisoning attack federated unlearning *** primary concept of SlideFU is to employ sliding window to construct the training process,where all operations are confined within the *** design a malicious detection scheme based on principal component analysis(PCA),which calculates the trust factors between compressed models in a low-cost way to eliminate unreliable *** confirming that the global model is under attack,the system activates the federated unlearning process,calibrates the gradients based on the updated direction of the calibration *** on two public datasets demonstrate that our scheme can recover a robust model with extremely high efficiency.
As the applications of large language models (LLMs) expand across diverse fields, their ability to adapt to ongoing changes in data, tasks, and user preferences becomes crucial. Traditional training methods with stati...
详细信息
Software developers and maintainers frequently conduct software refactorings to improve software quality. Identifying the conducted software refactorings may significantly facilitate the comprehension of software evol...
详细信息
Fault diagnosis technology is a method for locating faulty processors in multiprocessor systems, and it plays a crucial role in ensuring system stability, security and reliability. A widely used approach in this techn...
详细信息
Data sparsity poses a significant challenge for recommendation systems, prompting the research of Cross-Domain Recommendation (CDR). CDR aims to leverage more user-item interaction information from source domains to i...
详细信息
This article designs a 14-bit successive approximation register analog-to-digital converter(SAR ADC).A novel digital bubble sorting calibration method is proposed and applied to eliminate the effect of capacitor mis...
详细信息
This article designs a 14-bit successive approximation register analog-to-digital converter(SAR ADC).A novel digital bubble sorting calibration method is proposed and applied to eliminate the effect of capacitor mismatch on the linearity of the SAR ADC. To reduce the number of capacitors, a hybrid architecture of a high 8-bit binary-weighted capacitor array and a low 6-bit resistor array is adopted by the digital-to-analog(DAC). The common-mode voltage VCM-based switching scheme is chosen to reduce the switching energy and area of the DAC. The time-domain comparator is employed to obtain lower power consumption. Sampling is performed through a gate voltage bootstrapped switch to reduce the nonlinear errors introduced when sampling the input signal. Moreover, the SAR logic and the whole calibration is totally implemented on-chip through digital integrated circuit(IC) tools such as design compiler, IC compiler, etc. Finally, a prototype is designed and implemented using 0.18 μm bipolar-complementary metal oxide semiconductor(CMOS)-double-diffused MOS 1.8 V CMOS technology. The measurement results show that the SAR ADC with on-chip bubble sorting calibration method achieves the signal-to-noise-and-distortion ratio of 69.75 dB and the spurious-free dynamic range of 83.77 dB.
暂无评论