检索结果-内蒙古大学图书馆

13th International Conference on Control, Automation and Information sciences, ICCAIS 2024

作者： Van Ma, Linh Hussain, Muhammad Ishfaq Yow, Kin-Choong Jeon, Moongu School of Electrical Engineering and Computer Science Gist Korea Republic of Large-scale Ai Research Group Korea Republic of University of Regina Faculty of Engineering and Applied Science Canada

ISBN: (纸本)9798331542047

The MS-GLMB filter offers a robust framework for tracking multiple objects through the use of multi-sensor data. Building on this, the MV-GLMB and MV-GLMB-AB filters enhance the MS-GLMB capabilities by employing cameras for 3D multi-sensor multi-object tracking, effectively addressing occlusions. However, both filters depend on overlapping fields of view from the cameras to combine complementary information. In this paper, we introduce an improved approach that integrates an additional sensor, such as LiDAR, into the MS-GLMB framework for 3D multi-object tracking. Specifically, we present a new LiDAR measurement model, along with a multi-camera and LiDAR multi-object measurement model. Our experimental results demonstrate a significant improvement in tracking performance compared to existing MS-GLMB-based methods. Importantly, our method eliminates the need for overlapping fields of view, broadening the applicability of the MS-GLMB filter. Our source code for nuScenes dataset is available at https://***/linh-gist/ms-glmb-nuScenes. © 2024 IEEE.

关键词： Object tracking

来源：评论

学校读者我要写书评

暂无评论

Towards Large-Scale Benchmark Dataset for Remote Sensing Object Detection on Battlefield

Towards Large-Scale Benchmark Dataset for Remote Sensing Obj...

引用

2024 IEEE International Conference on Consumer Electronics-Asia, ICCE-Asia 2024

作者： Kim, Yechan Park, JongHyun Kim, Sihyun Kim, Sungheon Kim, SooYeon Ko, YeongMin Oh, Junggyun Jeon, Moongu School of Electrical Engineering and Computer Science GIST Korea Republic of Kyungpook National University Department of Robot and Smart System Engineering Korea Republic of HDC Labs Korea Republic of

ISBN: (纸本)9798331530839

Precise object detection allows military personnel to clearly understand their surroundings, leading to planning effective military strategies. Particularly, satellites and drones allow real-time surveillance over large areas, which is crucial for military operations. Collaboration among military, academic, and industrial institutions is required to promote innovation in military target detection. However, potential security concerns usually restrict access to real data from the military. To mitigate this issue, this paper proposes a novel approach to generate synthetic data for remote sensing object detection on the battlefield with ARMA3, one of the renowned military tactic games. With our method, the data for model training can be easily generated without any manpower. To demonstrate the efficacy of our approach, we provide a detailed analysis of the examples from our method. As ARMA3 is well-known for its realistic military combat simulation, we believe our method can effectively contribute to military object detection in remote sensing. © 2024 IEEE.

关键词： Military satellites

来源：评论

学校读者我要写书评

暂无评论

Reasoning Abilities of Large Language Models: In-Depth Analysis on the Abstraction and Reasoning Corpus

arXiv

引用

arXiv 2024年

作者： Lee, Seungpil Sim, Woochang Shin, Donghyeon Seo, Wongyu Park, Jiwon Lee, Seokki Hwang, Sanha Kim, Sejin Kim, Sundong Electrical Engineering and Computer Science GIST Korea Republic of AI Graduate School GIST Korea Republic of

The existing methods for evaluating the inference abilities of Large Language Models (LLMs) have been predominantly results-centric, making it challenging to assess the inference process comprehensively. We introduce a novel approach using the Abstraction and Reasoning Corpus (ARC) benchmark to evaluate the inference and contextual understanding abilities of LLMs in a process-centric manner, focusing on three key components from the Language of Thought Hypothesis (LoTH): Logical Coherence, Compositionality, and Productivity. Our carefully designed experiments reveal that while LLMs demonstrate some inference capabilities, they still significantly lag behind human-level reasoning in these three aspects. The main contribution of this paper lies in introducing the LoTH perspective, which provides a method for evaluating the reasoning process that conventional results-oriented approaches fail to capture, thereby offering new insights into the development of human-level reasoning in artificial intelligence systems. © 2024, CC BY.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

electrically Programable Tamm Plasmon for Broadband Optical Neuromorphic Computing

Electrically Programable Tamm Plasmon for Broadband Optical ...

引用

IEEE/LEOS International Conference on Optical MEMS

作者： Joo Hwan Ko Dong Hyun Seo Se Yeon Kim Young Min Song Electrical Engineering and Computer Science Gwangju Institute of Science and Technology (GIST) Gwangju Republic of Korea

ISBN: (数字)9798350384925

ISBN: (纸本)9798350384932

Inherent limitations of traditional computer architectures have catalyzed the development of brain-inspired hardware, with optical solutions increasingly favored for their energy efficiency, speed, and scalability. Accurately simulating the synaptic functions of neurons with optical components necessitates precise tuning of resonators to specific wavelengths for multiplexed operations within photonic integrated circuits (PICs), a process integral to enabling massively parallel signal processing. However, as integration density increases, the task becomes further complicated by the challenge of arranging resonators of varying wavelengths within the confined two-dimensional space of a waveguide. Here, we propose a broadband neuromorphic computing approach using free-space optics that simplifies wavelength selection through straightforward parameter adjustments, consequently facilitating superior spatial configurations in a three-dimensional layout. We have implemented selective optical resonators enhanced by Tamm plasmon structures, which produce strong resonance through precise impedance matching, even in non-sophisticated planar structures, by integrating with poly(3,4-ethylenedioxythiophene):polystyrene sulfonate (PEDOT:PSS). By modulating the thickness of the Tamm plasmon structures, we easily modified the resonance frequency with high precision. Consequently, leveraging the memory characteristics and non-volatile properties of PEDOT:PSS, we achieved expected results in enhancing the computational capabilities and efficiency of our system with a broadband and multiplexed optical signals.

关键词： Integrated optics Neuromorphic engineering Optical resonators Plasmons Resonance Broadband communication Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Low-Powered, Full-Color Reflective Display based on Electrochromic Resonator

Low-Powered, Full-Color Reflective Display based on Electroc...

引用

IEEE/LEOS International Conference on Optical MEMS

作者： Hyo Eun Jeong Joo Hwan Ko Young Min Song Electrical Engineering and Computer Science Gwangju Institute of Science and Technology (GIST) Gwangju Republic of Korea

ISBN: (数字)9798350384925

ISBN: (纸本)9798350384932

Dynamically tunable reflective display is attractive for their applicability, due to high energy efficiency and clear outdoor visibility. While many reflective displays have been demonstrated, its monochromic color and high operating voltage are remained challenges and restricted their use in diverse electronics. Here, we introduce the full-colored monopixel reflective display (FMRD), which is achieved through a dual cavity structure with active material and lossy medium. As an active layer, polyaniline (PANI) forms a resonant cavity capable of changing its optical properties within a voltage range of -0.2 V to 0.8 V. To obtain a clear reflective color, we enhance light-matter interaction through impedance- matching between lossy medium and active layer. By designing porosity of lossy material with glancing angle deposition, it can be obtained and exhibits a deep reflectance dip. Additionally, the device successfully achieves the generation of the full red-green-blue (RGB) colors within a single pixel. FMRD is maintained half a day without power supply and reactivated its vivid colors after 4 months because self-passivation layer prevents the germanium oxidation. The protonation-based redox reaction plays a crucial role in achieving a fast response, with a rising time of 34 ms and a falling time of 171 ms. The photonic resonator operates at only 0.42 mW/cm 2 per cycle and demonstrates scalability by addressing 25 pixels in a 5×5 matrix, indicating its applicability for larger-scale metrics.

关键词： Optical losses Visualization Stability criteria Optical resonators Color Optical variables control Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

Depth Prompting for Sensor-Agnostic Depth Estimation

Depth Prompting for Sensor-Agnostic Depth Estimation

引用

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Jin-Hwi Park Chanhwi Jeong Junoh Lee Hae-Gon Jeon AI Graduate School GIST School of Electrical Engineering and Computer Science South Korea

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Dense depth maps have been used as a key element of visual perception tasks. There have been tremendous efforts to enhance the depth quality, ranging from optimization-based to learning-based methods. Despite the remarkable progress for a long time, their applicability in the real world is limited due to systematic measurement biases such as density, sensing pattern, and scan range. It is well-known that the biases make it difficult for these methods to achieve their generalization. We observe that learning a joint representation for input modalities (e.g., images and depth), which most recent methods adopt, is sensitive to the biases. In this work, we disentangle those modalities to mitigate the biases with prompt engineering. For this, we design a novel depth prompt module to allow the desirable feature representation according to new depth distributions from either sensor types or scene configurations. Our depth prompt can be embedded into foundation models for monocular depth estimation. Through this embedding process, our method helps the pretrained model to be free from restraint of depth scan range and to provide absolute scale depth maps. We demonstrate the effectiveness of our method through extensive evaluations. Source code is publicly available at https://***/JinhwiPark/DepthPrompting.

关键词： Learning systems Systematics Source coding Estimation Time measurement Stability analysis Sensors

来源：评论

学校读者我要写书评

暂无评论

Input Impedance Reduction Techniques for the Resistive Dipoles on Printed Circuit Board

Input Impedance Reduction Techniques for the Resistive Dipol...

引用

International Symposium on Antennas and Propagation (ISAP)

作者： Jaeuk Choi Kangwook Kim School of Electrical Engineering and Computer Science Gwangju Institute of Science and Technology (GIST) Gwangju Republic of Korea

ISBN: (数字)9798350364774

ISBN: (纸本)9798350364781

This study investigated methods to reduce the high input impedance of a resistive dipole antenna with the Wu-King profile. The resistive dipole antenna has a high reflection coefficient due to its high input impedance, which is largely dependent on the permittivity of the PCB substrate and the width of the antenna arms. We designed a resistive dipole with a resistance per unit length of R 0 = 3337.5 Ω/m. We show that for the resistive dipole with a width of 3.2 mm on a PCB with a dielectric constant of ϵ r = 4.4, the input impedance is improved by approximately 165 Ω compared to the case with ϵ r = 1.0 and a width of 2.0 mm.

关键词： Resistance Dipole antennas Printed circuits Reflector antennas Reflection Reflection coefficient Impedance Permittivity Substrates Loaded antennas

来源：评论

学校读者我要写书评

暂无评论

Artificial compound eye camera for wide field of view and high-resolution imaging

Artificial compound eye camera for wide field of view and hi...

引用

IEEE International Conference on Flexible and Printable Sensors and Systems (FLEPS)

作者： Sehui Chang Young Min Song Gwangju Institute of Science and Technology (GIST): School of Electrical Engineering and Computer Science Gwangju Korea GIST: AI Graduate School Gwangju Korea GIST: Department of Semiconductor Engineering Gwangju Korea

ISBN: (数字)9798350383263

ISBN: (纸本)9798350383270

Insect compound eyes have been garnered a great attention in the fields of optical design, owing to their remarkably economical, but sophisticated visual structures and process. However, a rigid and planar imaging sensor hardly supporting curved micro-optics hinders full implementation of insect vision system. Here, we present a bioinspired artificial compound eye camera for wide field of view and high-resolution imaging based on the hemispherical microlens array and tapered imaging fiber bundles. The hemispherical microlens array focuses a wide field of view image on the curved facet of tapered fiber bundle, which relays the image to another facet contacted on commercial image sensor providing high resolution. The integrated compound eye module suggests a facile way to fill the technical gap for practical applications of artificial insect vision systems.

关键词： Optical fiber sensors Visualization Machine vision Insects High-resolution imaging Cameras Compounds

来源：评论

学校读者我要写书评

暂无评论

Optical Neuromorphic modulators based on Tamm Plasmon Resonators with PEDOT:PSS

Optical Neuromorphic modulators based on Tamm Plasmon Resona...

引用

IEEE International Conference on Flexible and Printable Sensors and Systems (FLEPS)

作者： Young Min Song Joo Hwan Ko Dong Hyun Seo Gwangju Institute of Science and Technology (GIST) School of Electrical Engineering and Computer Science Gwangju Korea GIST: AI Graduate School Gwangju Korea GIST: Department of Semiconductor Engineering Gwangju Korea

ISBN: (数字)9798350383263

ISBN: (纸本)9798350383270

Reconfigurable optical devices are increasingly recognized for their potential in enhancing high-density optical interconnects, photonic switching, and memory systems. While various optical modulators utilizing active materials have been successfully developed, achieving high modulation depth with low operating voltage in the near-infrared (NIR) spectrum remains a significant challenge. This spectrum is particularly important for free-space communication and imaging technologies. In this context, an electrically switchable device employing Tamm plasmon coupled with poly(3,4-ethylenedioxythiophene): polystyrene sulfonate (PEDOT:PSS) has been introduced. This innovative approach enables substantial modulation depth throughout the NIR spectrum, effectively absorbing incident light even under epsilon near zero conditions. The device boasts an impressive optical modulation efficiency of over 88% at a CMOS-compatible voltage of ±1 V, achieved through meticulous electrical manipulation of charge carrier density via an electrochemical doping/dedoping process. Moreover, this device opens new avenues for creating non-volatile optical memory storage solutions, exhibiting rewritable multi-memory states and demonstrating long-term potentiation/depression capabilities, akin to neuromorphic behavior.

关键词： Neuromorphics Optical switches Optical resonators Voltage Plasmons Optical imaging Sensor systems

来源：评论

学校读者我要写书评

暂无评论

Programmable Plasmonic-Nanopixels for High- Density Chromatic Information Encryption

Programmable Plasmonic-Nanopixels for High- Density Chromati...

引用

IEEE/LEOS International Conference on Optical MEMS

作者： Ji-Eun Yeo Joo Hwan Ko Young Min Song School of Electrical Engineering and Computer Science Gwangju Institute of Science and Technology (GIST) Gwangju Republic of Korea

ISBN: (数字)9798350384925

ISBN: (纸本)9798350384932

Efficiently modulating light-matter interactions in visible wavelength range holds significant importance for various applications such as optical encryption, switching mechanisms, and structural color displays. In this study, we introduce a novel approach for achieving electrochromic nanopixels by combining polyaniline (PANI) with metal-coated Si nanowires (Si NWs). Our optical modulation technique enables operation within an expanded color spectrum at a remarkably low voltage level below 1V, thereby ensuring energy-efficient performance. The intrinsic optical properties transition of PANI from reduced (dielectric) to oxidized (lossy) form, resulting in significant changes in its complex refractive index. Oxidation of PANI induces the effective cancellation of surface plasmon polaritons (SPPs) while simultaneously leading to strong broadband absorption due to an increased extinction coefficient. The fabrication process is versatile and can be tailored to meet various optical information encryption needs, scalable up to 8-inch wafers. A unique chromatic modulation capability enables the data encryption while leveraging the angle dependency of the waveguide to amplify the information capacity to 10 bits per unit cell. To further enhance the data storage capacity, our proposed system can incorporate multiple resonant structures based on diverse nanowire domains with varying dimensions to enhance chromatic changes.

关键词： Image color analysis Optical device fabrication Optical variables control Optical coupling Encryption Nanowires Optical modulation

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：