| Nama Pengarang | : | Yang Ji; Ying Sun; Yuting Zhang; Zhigaoyuan Wang |
| Nama Majalah/Jurnal | : | Proceedings of the IEEE |
| Volume / Nomor Edisi | : | 113 (No. 8) |
| Nomor Halaman | : | 783-813 |
| Abstraksi Narasi | : | Neural networks have achieved remarkable success across various fields. However, the lack of interpretability limits their practical use, particularly in critical decision-making scenarios. Posthoc interpretability, which provides explanations for pretrained models, is often at risk of fidelity and robustness. This has inspired a rising interest in self-interpretable neural networks (SINNs), which inherently reveal the prediction rationale through model structures. Despite this progress, existing research remains fragmented, relying on intuitive designs tailored to specific tasks. To bridge these efforts and foster a unified framework, we first collect and review existing works on SINNs and provide a structured summary of their methodologies from five key perspectives: attribution-based, function-based, concept-based, prototype-based, and rule-based self-interpretation. We also present concrete, visualized examples of model explanations and discuss their applicability across diverse scenarios, including image, text, graph data, and deep reinforcement learning (DRL). Additionally, we summarize existing evaluation metrics for self-interpretation and identify open challenges in this field, offering insights for future research. To support ongoing developments, we present a publicly accessible resource to track advancements in this domain: https://github.com/yangji721/Awesome-Self-Interpretable-Neural-Network |
| Nama Pengarang | : | Zecai Lin; Cheng Zhou; Shaoping Huang |
| Nama Majalah/Jurnal | : | Proceedings of the IEEE |
| Volume / Nomor Edisi | : | 113 (No. 8) |
| Nomor Halaman | : | 752-782 |
| Abstraksi Narasi | : | Deformation of flexible robots can be practically assessed using extension/compression, shear, curvature, and torsion. Sensing based on one or more of the above characteristics enables closed-loop control for delicate tasks that require precision and dexterity. Due to the increasing popularity of flexible robotics in recent years, significant research effort has been directed to this burgeoning field. Although numerous studies have addressed soft sensing technologies, their successful integration into flexible robotic systems remains limited. This article provides a comprehensive review of sensing methods, from multidimensional deformation to the underlying principles of deriving hard-to-measure deformation from surrogate parameters. It focuses on sensing modalities such as strain measurement via piezoelectric, capacitive, resistive, and optical techniques. The applications of deformation sensing in industrial and service robotics are described. Future challenges and potential research issues including resolution, conformability, multifunctionality, crosstalk, and miniaturization are discussed. The need for a synergistic approach across disciplines is highlighted, emphasizing the integration of new materials, microstructures, advanced manufacturing technologies, and state-of-the-art signal processing techniques. |
| Nama Pengarang | : | Luís Crespo; Nuno Neves |
| Nama Majalah/Jurnal | : | Proceedings of the IEEE |
| Volume / Nomor Edisi | : | 113 (No. 8) |
| Nomor Halaman | : | 713-751 |
| Abstraksi Narasi | : | In the past few years, there has been a renewed effort to advance general-purpose architectures. In particular, to deliver performance and energy efficiency advantages, several techniques have been applied based on new forms of specialization while maintaining usability. As a result, data movement and communication have become the primary bottlenecks in computer systems. To overcome this, one of the most recent breakthroughs has been the introduction of data streaming mechanisms, just like those used in accelerators, into modern general-purpose processors (GPPs). This article comprehensively reviews stream-based architectures, tracing their development from accelerator solutions to their recent adoption in GPPs. This survey starts by introducing the fundamental principles of stream specialization, followed by a taxonomy for memory accesses, and formal mathematical models to represent them as data streams. Then, it categorizes different topologies of data stream specialization and examines them from a compiler’s perspective. Some of the most representative architectures proposed in the past few years, including instruction set architecture (ISA) and streaming engines, are described, followed by a comparative analysis that highlights their key features and presents quantitative evaluations. Then, we discuss some open challenges and suggest directions for future research in stream-based architectures. |
| Nama Pengarang | : | Peter M. Grant; John S. Thompson; Simon Watts |
| Nama Majalah/Jurnal | : | Proceedings of the IEEE |
| Volume / Nomor Edisi | : | 113 (No. 7) |
| Nomor Halaman | : | 693-706 |
| Abstraksi Narasi | : | In February 1935, the Daventry radar experiment, led by Robert Watson-Watt, detected the presence of an aircraft by using reflected radio waves [1]. Later that year, the Chain Home radar system design started [1], with the first installation in 1938 covering the approaches to the Thames Estuary. This enabled the detection of incoming aircraft at 160-km range and directed interception fighters to within 6.5 km of the target. This system used 100- kW–1-MW high-energy pulses generated at 20–50 MHz and was later extended to cover the U.K. east and south coasts as well as parts of the west coast. Compared to today’s microwave radar systems, this was a basic design with wooden and steel towers supporting the physically large high-frequency (HF) antennas. |
| Nama Pengarang | : | Tom Backstrom |
| Nama Majalah/Jurnal | : | Proceedings of the IEEE |
| Volume / Nomor Edisi | : | 113 (No. 7) |
| Nomor Halaman | : | 668-692 |
| Abstraksi Narasi | : | Speech technology for communication, accessing information, and services has rapidly improved in quality. It is convenient and appealing because speech is the primary mode of communication for humans. Such technology, however, also presents proven threats to privacy. Speech is a tool for communication, and thus, it will inherently contain private information. Importantly, it also contains a wealth of side information, including details about health, emotions, affiliations, and relationships, all of which are private. Exposing such private information can lead to serious threats such as price gouging, harassment, extortion, and stalking. This article is a tutorial on privacy issues related to speech technology, modeling their threats, approaches for protecting users’ privacy, measuring the performance of privacy-protecting methods, perception of privacy, as well as societal and legal consequences. In addition to a tutorial overview, it also presents lines for further development where improvements are most urgently needed. |
| Nama Pengarang | : | Swaroop Ghosh; Suryansh Upadhyay; Abdullah Ash Saki |
| Nama Majalah/Jurnal | : | Proceedings of the IEEE |
| Volume / Nomor Edisi | : | 113 (No. 7) |
| Nomor Halaman | : | 640-667 |
| Abstraksi Narasi | : | Quantum computing (QC) is an emerging paradigm with the potential to transform numerous application domains by addressing classically intractable problems. However, its growing presence in cyberspace has introduced new security and privacy challenges. Similar to classical computing systems, the QC stack including software and hardware relies extensively on third parties, many of which are emerging and trust-seeking or less-trusted. This stack often contains sensitive intellectual property (IP) that demands protection. Unique features of quantum systems can enable classical-style attacks: for instance, crosstalk in multitenant settings can facilitate fault-injection attacks, while malicious calibration services can misreport error rates or miscalibrate qubits to induce denial-of-service (DoS) conditions. Given the high cost and limited availability of likely trustworthy quantum hardware, users may be enticed to explore emerging and trust-seeking but cheaper and readily available quantum hardware, which can enable the stealth of IP and tampering of quantum programs and/or computation outcomes. Similarly, emerging compilation services may compromise circuit confidentiality or insert Trojans. Despite the strategic significance of QC and its potential to process sensitive information, its security and privacy concerns remain underexplored. This article presents a comprehensive overview of QC fundamentals, key vulnerabilities, recent attack vectors, and corresponding defenses, and concludes with directions for future research to strengthen the quantum security community. |
| Nama Pengarang | : | Andrew Boutros; Aman Arora; Vaughn Betz |
| Nama Majalah/Jurnal | : | Proceedings of the IEEE |
| Volume / Nomor Edisi | : | 113 (No. 7) |
| Nomor Halaman | : | 613-639 |
| Abstraksi Narasi | : | Deep learning (DL) is becoming the cornerstone of numerous applications both in large-scale datacenters and at the edge. Specialized hardware is often necessary to meet the performance requirements of state-of-the-art DL models, but the rapid pace of change in DL models and the wide variety of systems integrating DL make it impossible to create custom computer chips for all but the largest markets. Field-programmable gate arrays (FPGAs) present a unique blend of reprogrammability and direct hardware execution that make them suitable for accelerating DL inference. They offer the ability to customize processing pipelines and memory hierarchies to achieve lower latency and higher energy efficiency compared to general-purpose central processing units (CPUs) and graphics processing units (GPUs), at a fraction of the development time and cost of custom chips. Their diverse and high-speed inputs/outputs (IOs) also enable directly interfacing the FPGA to the network and/or a variety of external sensors, making them suitable for both datacenter and edge use cases. As DL has become an ever more important workload, FPGA architectures are evolving to enable higher DL performance. In this article, we survey both academic and industrial FPGA chip architecture enhancements for DL. First, we give a brief introduction on the basics of FPGA architecture and how its components lead to strengths and weaknesses for DL applications. Next, we discuss different design styles of DL inference accelerators implemented on FPGAs that achieve state-of-the-art performance and productive development flows, ranging from model-specific dataflow styles to software-programmable overlay styles. We survey DL-specific enhancements to traditional FPGA building blocks including the logic blocks (LBs), arithmetic circuitry, and on-chip memories, as well as new DL-specialized blocks that integrate into the FPGA fabric to accelerate tensor computations. Finally, we discuss hybrid devices that combine processors and coarse-grained accelerator blocks with FPGA-like interconnect and networks-on-chip (NoCs), and highlight promising future research directions. |
| Nama Pengarang | : | - |
| Nama Majalah/Jurnal | : | IEEE Computational intelligence |
| Volume / Nomor Edisi | : | 12 (No. 2) |
| Nomor Halaman | : | 56-69 |
| Abstraksi Narasi | : | It is difficult to implement optimal control for a system whose model is unknown and operation environment is uncertain, such as the intelligent cruise control of vehicles. This article will address the problem from the perspective of reinforcement learning by learning the optimal policy from the state transition data. The model-free optimal control algorithm is employed to approximate the optimal control policy for the intelligent cruise control system, which considers the comfort performance and the safety performance comprehensively by setting up a total performance index. The algorithm is implemented by two multi-layer neural networks which are the critic network and the actor network. The critic and actor networks are employed to approximate the stateaction value function and the control action, respectively. In addition, a data collecting strategy is proposed to obtain the state transition data distributed uniformly in the state action space from the running trajectory of the host car. The critic network and the action network are trained alternatively by the collected data until converging. The convergent action network is used to obtain the optimal control policy. At last, the policy is tested on a hardware-in-the-loop simulator built upon dSPACE by comparing with a linear quadratic regulator (LQR) controller and a proportion integration differentiation (PID) controller. Results show its excellent performance on both aspects of the safety and the comfort. |
| Nama Pengarang | : | Fangqing Liu,Xiaoyan Zhuo,Han Huang,Zhifeng Hao |
| Nama Majalah/Jurnal | : | IEEE Computational intelligence |
| Volume / Nomor Edisi | : | 12 (No. 2) |
| Nomor Halaman | : | 46-55 |
| Abstraksi Narasi | : | One of the challenges of search-based software testing is the automated test case generation for path coverage (ATCG-PC). There have been several meta-heuristic algorithms proposed to cope with this challenge. However, the cost of test cases generated by the algorithms is still high and even infinite for complete path coverage. The goal of the current study is to propose an improved meta-heuristic algorithm for ATCGPC. Unlike the previous research, our study focuses on how to use more heuristic information to generate more uncovered paths with fewer test cases. We combine a self-adaptive fitness function and a modified differential evolution algorithm to implement this plan. Follow-up experimental studies on eight benchmark problems of classical programs show that the proposed algorithm outperforms those solutions obtained by immune genetic algorithm, artificial bee colony, and random search. |
| Nama Pengarang | : | - |
| Nama Majalah/Jurnal | : | Analisa |
| Volume / Nomor Edisi | : | X-4, APRIL (No. 4) |
| Nomor Halaman | : | 337-361 |
| Abstraksi Narasi | : | - |