Accurate prediction of thermodynamic properties is essential in drug discovery and materials science. Molecular dynamics (MD) simulations provide a principled approach to this task, yet they typically rely on prohibit...
详细信息
This study examines to what extent the testing of traditional software components and machine learning (ML) models fundamentally differs or not. While some researchers argue that ML software requires new concepts and ...
详细信息
This study examines to what extent the testing of traditional software components and machine learning (ML) models fundamentally differs or not. While some researchers argue that ML software requires new concepts and perspectives for testing, our analysis highlights that, at a fundamental level, the specification and testing of a software component are not dependent on the development process used or on implementation details. Although the software engineering/computerscience (SE/CS) and Data science/ML (DS/ML) communities have developed different expectations, unique perspectives, and varying testing methods, they share clear commonalities that can be leveraged. We argue that both areas can learn from each other, and a non-dual perspective could provide novel insights not only for testing ML but also for testing traditional software. Therefore, we call upon researchers from both communities to collaborate more closely and develop testing methods and tools that can address both traditional and ML software components. While acknowledging their differences has merits, we believe there is great potential in working on unified methods and tools that can address both types of software.
We monitor a 524km live network link using an FPGA-based sensing-capable coherent transceiver prototype during a human-caused cable break. Post-analysis of polarization data reveals minute-level potential warning prec...
详细信息
In pursuit of autonomous vehicles, achieving human-like driving behavior is vital. This study introduces adaptive autopilot (AA), a unique framework utilizing constrained-deep reinforcement learning (C-DRL). AA aims t...
详细信息
ISBN:
(数字)9798331505929
ISBN:
(纸本)9798331505936
In pursuit of autonomous vehicles, achieving human-like driving behavior is vital. This study introduces adaptive autopilot (AA), a unique framework utilizing constrained-deep reinforcement learning (C-DRL). AA aims to safely emulate human driving to reduce the necessity for driver intervention. Focusing on the car-following scenario, the process involves (i) extracting data from the highD natural driving study and categorizing it into three driving styles using a rule-based classifier; (ii) employing deep neural network (DNN) regressors to predict human-like acceleration across styles; and (iii) using C-DRL, specifically the soft actor-critic Lagrangian technique, to learn human-like safe driving policies. Results indicate effectiveness in each step, with the rule-based classifier distinguishing driving styles, the regressor model accurately predicting acceleration, outperforming traditional car-following models, and C-DRL agents learning optimal policies for humanlike driving across styles.
To have competent 3D lane detection for real-world driving, a massive amount of data from all over the world is needed, but data collection and manual annotation are costly and time-consuming. The diversity of data co...
To have competent 3D lane detection for real-world driving, a massive amount of data from all over the world is needed, but data collection and manual annotation are costly and time-consuming. The diversity of data collected by developmental cars might still be limited compared to the data collected by a large fleet of customer cars. Federated learning enables training models on edge without transferring data out of devices. However, training supervised learning tasks at the edge is directly tied to having access to high-quality labels, which is limited at the *** this paper, we propose a fully automatic method to generate 3D lane labels at the edge using a pre-recorded HD map to enable the federated training of the 3D lane detection model. As a reference, a semi-automatic method is applied for creating a 3D-lane dataset used as ground truth. Our experimental results show that the model can achieve comparable performance when training on the same dataset in both a centralized and a decentralized manner. And the models trained on semi-automatic labeled datasets slightly outperform those trained on fully-automatically labeled datasets. This study shows that a well-performing 3D lane detection model can be trained in a supervised and fully decentralized manner, and most importantly, data privacy at the edge is guaranteed.
Future wireless networks will integrate sensing, learning and communication to provide new services beyond communication and to become more resilient. Sensors at the network infrastructure, sensors on the user equipme...
详细信息
Single-vehicle accidents are the most common type of fatal accidents in Sweden, where a car drives off the road and runs into hazardous roadside objects. Proper installation and maintenance of protective objects, such...
详细信息
In this paper we consider the stacking of isotonic regression and the method of rearrangement with the empirical estimator to estimate a discrete distribution with an infinite support. The estimators are proved to be ...
详细信息
We perform exploratory ASIC design of key DSP and FEC units for 400-Gbit/s coherent data-center interconnect receivers. In 22-nm CMOS, the considered units together dissipate 5W, suggesting implementation feasibility ...
详细信息
We review different real-Time implementation platforms for digital signal processing. We discuss circuit implementation of coherent receivers and design trade-offs in-volving circuit complexity, throughput and power d...
详细信息
暂无评论