检索结果-内蒙古大学图书馆

Progress in Electromagnetic Research Symposium (PIERS)

作者： Mengying Dong Xianliang Wu Yue Ma Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University Hefei China

ISBN: (纸本)9781665409889

Aiming at the single characteristic of traditional antenna bandwidth and working polarization state, a broadband circularly polarized antenna based on metamaterial surface is proposed. The antenna is composed of a rectangular microstrip antenna and an upper metasurface. The distance between the two is h. The metasurface is composed of $5\times 5$ square metal elements with equiangular structures on both sides. The linearly polarized rectangular microstrip antenna as the feeding antenna radiates linearly polarized electromagnetic waves, and the metasurface converts the linearly polarized electromagnetic waves radiated by the feeding antenna into circularly polarized electromagnetic waves. This paper analyzes the principle of circular polarization after the antenna is loaded with a metasurface. The simulation software HFSS is used to perform parameter scanning and simulation analysis on the antenna with the size of 0.88$\lambda\times 0.88$$\lambda\times 0.04$$\lambda \mathrm{m}^{3}$ and the center frequency of 8.5 GHz. The results show that the impedance bandwidth of the atenna S 11

关键词： Polarization Microstrip antennas Bandwidth Metasurfaces Antenna feeds Broadband antennas Microstrip

来源：评论

学校读者我要写书评

暂无评论

Highly Selective UWB Bandpass Filter with Dual Notch Bands Using Stub Loaded Multiple-mode Resonator

Highly Selective UWB Bandpass Filter with Dual Notch Bands U...

引用

Progress in Electromagnetic Research Symposium (PIERS)

作者： Guangxiu Zhao Minquan Li Ran Zhao Ziyun Tu Yajing Yan Xiaming Mo Key Laboratory of Intelligent Computing & Signal Processing Ministry of Education Anhui University Hefei China

ISBN: (纸本)9781665409889

In this paper, a low-loss ultra-wideband (UWB) bandpass filter with high selectivity and multiple notch bands is proposed. The filter is based on multi-mode resonator (MMR) which mainly consists of a step impedance transmission line that excites several resonant modes. The MMR structure is loaded a T-shaped open-circuited stub which generate two transmission zeros at the edge of the passband to improve selectivity. Meanwhile, MMR structure is also loaded two pairs of short high impedance stubs to extend the upper stopband of the filter. In addition, to address the in-band interference issues, by coupling a tunable parasitic unit to generate dual band notches at 5.8GHz (WLAN) and 8.1 GHz (X-band). Simulated result indicates that the designed filter’s passband is from 3.18 GHz to 10.67 GHz, the insertion loss is lower than 0.5 dB and its 3 dB relative bandwidth is 108.1%. At the same time, the filter has a high selectivity (30dB skirt factor is 0.917) and a stopband of up to 20 GHz (S 21 < −15dB).

关键词： Band-pass filters Couplings Wireless LAN Resonator filters Insertion loss Bandwidth Filtering theory

来源：评论

学校读者我要写书评

暂无评论

A Novel Design of Wide Stopband Filter with High Selectivity

A Novel Design of Wide Stopband Filter with High Selectivity

引用

Progress in Electromagnetic Research Symposium (PIERS)

作者： Minquan Li Xiaming Mo Ran Zhao Yajing Yan Guangxiu Zhao Ziyun Tu Key Laboratory of Intelligent Computing & Signal Processing Ministry of Education Anhui University Hefei China

ISBN: (纸本)9781665409889

In this paper, a wide stopband bandpass filter with high selectivity is designed, which uses three stepped impedance resonators to form the main body of the filter, which improves the stopband width of the filter, and has a wide stopband to 3f 0 . The parasitic passband of the filter is suppressed by etching the double T-shaped defected ground structure that changes the current distribution on the floor, generate a band gap with band-stop characteristics, the stopband width is over 4f 0 . Transmission zero is introduced by cross coupling resonance. There are two coupling paths for the signal from the input to output, one transmission zero is obtained on the upper stopband, then the selectivity of the filter is improved obviously. The simulation results show that the return loss of the filter can reach 46 dB and the insertion loss is less than 0.5 dB in the passband of 4.8 GHz-5.0 GHz. The stopband attenuation of 5.8 GHz-20 GHz is better than 20 dB. The rectangular coefficient of 30 dB/3dB is 3.72, and the size is 12 * 45.9 mm 2 .

关键词： Band-pass filters Couplings Photonic band gap Simulation Resonator filters Insertion loss Attenuation

来源：评论

学校读者我要写书评

暂无评论

A novel miniaturized l-band filter with great stopband characteristics using interdigitated coupled lines CRLH-TL structure

引用

Progress In Electromagnetics Research C 2021年 114卷 57-67页

作者： Wang, Peng Duan, Kaiyue Li, Minquan Zhang, Man Jin, Baokun Anqing Normal University Anqing246011 China Key Lab of Ministry of Education of Intelligent Computing & Signal Processing Anhui University Hefei230601 China

This paper proposes a novel bandpass filter for L-band based on CRLH TL, which is mainly formed by coupling a high-pass characteristic module with a low-pass characteristic module in a cascade. The high-pass module consists of an interdigitated coupled line and a grounding via, owning to its singular characteristics, which the miniaturization is realized. The low-pass module is composed of a C-type resonator with high-low impedance lines, which can realize great sideband attenuation characteristics. To further improve its out-of-band rejection characteristics, a complementary split-ring resonator (CSRR) defective ground structure with single-pole attenuation characteristics is loaded, and a transmission zero is introduced at 2.5f0 out-of-band. The test results are in great agreement with the simulation ones, and the dimensions are only 0.20λg * 0.22λg. Compared with other similar types, the filter proposed in this paper has miniaturization, great passband selection characteristics, stopband characteristics, and the advantage of low insertion loss. © 2021, Electromagnetics Academy. All rights reserved.

关键词： Bandpass filters

来源：评论

学校读者我要写书评

暂无评论

Rethink Video Retrieval Representation for Video Captioning

SSRN

引用

SSRN 2024年

作者： Tian, Mingkai Li, Guorong Qi, Yuankai Wang, Shuhui Sheng, Quan Z. Huang, Qingming School of Computer Science and Technology Key Lab of Big Data Mining and Knowledge Management University of Chinese Academy of Sciences China School of Computing Macquarie University Australia Key Laboratory of Intelligent Information Processing Institute of Computer Technology Chinese Academy of Sciences China

Video captioning, a challenging task targeting the automatic generation of accurate and comprehensive descriptions based on video content, has witnessed substantial success recently driven by bridging video representations and textual semantics. Inspired by the nature of the video retrieval task, which learns visual features strongly related to text queries, we propose to take advantage of visual representation learning from the video retrieval framework to tackle video captioning tasks. However, a simple direct application of video retrieval models performs poorly due to the weak ability to capture sufficient video details and temporal information required for video captioning. To increase the attention on details, we propose a learnable token shift module, which flexibly captures subtle movements in local regions across the temporal sequence. Furthermore, we devise a Refineformer, which learns to integrate local video tokens strongly related to desired captions via a cross-attention mechanism. Extensive experiments on MSVD, MSR-VTT and VATEX demonstrate the favorable performance of our method. © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Synthesize, Diagnose, and Optimize: Towards Fine-Grained Vision-Language Understanding

arXiv

引用

arXiv 2023年

作者： Peng, Wujian Xie, Sicheng You, Zuyao Lan, Shiyi Wu, Zuxuan Shanghai Key Lab of Intell. Info. Processing School of CS Fudan University China Shanghai Collaborative Innovation Center of Intelligent Visual Computing China NVIDIA United States

Vision language models (VLM) have demonstrated remarkable performance across various downstream tasks. However, understanding fine-grained visual-linguistic concepts, such as attributes and inter-object relationships, remains a significant challenge. While several benchmarks aim to evaluate VLMs in finer granularity, their primary focus remains on the linguistic aspect, neglecting the visual dimension. Here, we highlight the importance of evaluating VLMs from both a textual and visual perspective. We introduce a progressive pipeline to synthesize images that vary in a specific attribute while ensuring consistency in all other aspects. Utilizing this data engine, we carefully design a benchmark, SPEC, to diagnose the comprehension of object size, position, existence, and count. Subsequently, we conduct a thorough evaluation of four leading VLMs on SPEC. Surprisingly, their performance is close to random guess, revealing significant limitations. With this in mind, we propose a simple yet effective approach to optimize VLMs in fine-grained understanding, achieving significant improvements on SPEC without compromising the zero-shot performance. Results on two additional fine-grained benchmarks also show consistent improvements, further validating the transferability of our approach. Code and data are available at https://***/wjpoom/SPEC. Copyright © 2023, The Authors. All rights reserved.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

OmniViD: A Generative Framework for Universal Video Understanding

OmniViD: A Generative Framework for Universal Video Understa...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Junke Wang Dongdong Chen Chong Luo Bo He Lu Yuan Zuxuan Wu Yu-Gang Jiang Shanghai Key Lab of Intell. Info. Processing School of CS Fudan University Shanghai Collaborative Innovation Center of Intelligent Visual Computing Microsoft Cloud + AI Microsoft Research Asia University of Maryland College Park

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

The core of video understanding tasks, such as recognition, captioning, and tracking, is to automatically de-tect objects or actions in a video and analyze their temporal evolution. Despite sharing a common goal, different tasks often rely on distinct model architectures and annotation formats. In contrast, natural language processing benefits from a unified output space, i.e., text sequences, which simplifies the training of powerful foundational language models, such as GPT-3, with extensive training cor-pora. Inspired by this, we seek to unify the output space of video understanding tasks by using languages as labels and additionally introducing time and box tokens. In this way, a variety of video tasks could be formulated as video-grounded token generation. This enables us to address var-ious types of video tasks, including classification (such as action recognition), captioning (covering clip captioning, video question answering, and dense video captioning), and localization tasks (such as visual object tracking) within a fully shared encoder-decoder architecture, following a generative framework. Through comprehensive experiments, we demonstrate such a simple and straightforward idea is quite effective and can achieve state-of-the-art or compet-itive results on seven video benchmarks, providing a novel perspective for more universal video understanding. Code is available at https://***/wangjk666/OmniVid.

关键词： Training Location awareness Visualization Vocabulary Computer architecture Benchmark testing Robustness

来源：评论

学校读者我要写书评

暂无评论

BEVNeXt: Reviving Dense BEV Frameworks for 3D Object Detection

arXiv

引用

arXiv 2023年

作者： Li, Zhenxin Lan, Shiyi Alvarez, Jose M. Wu, Zuxuan Shanghai Key Lab of Intell. Info. Processing School of CS Fudan University China Shanghai Collaborative Innovation Center of Intelligent Visual Computing China NVIDIA United States

Recently, the rise of query-based Transformer decoders is reshaping camera-based 3D object detection. These query-based decoders are surpassing the traditional dense BEV (Bird’s Eye View)-based methods. However, we argue that dense BEV frameworks remain important due to their outstanding abilities in depth estimation and object localization, depicting 3D scenes accurately and comprehensively. This paper aims to address the drawbacks of the existing dense BEV-based 3D object detectors by introducing our proposed enhanced components, including a CRF-modulated depth estimation module enforcing object-level consistencies, a long-term temporal aggregation module with extended receptive fields, and a two-stage object decoder combining perspective techniques with CRF-modulated depth embedding. These enhancements lead to a "modernized" dense BEV framework dubbed BEVNeXt. On the nuScenes benchmark, BEVNeXt outperforms both BEV-based and query-based frameworks under various settings, achieving a state-of-the-art result of 64.2 NDS on the nuScenes test set. Code will be available at https://***/woxihuanjiangguo/BEVNeXt. Copyright © 2023, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

GridMM: Grid Memory Map for Vision-and-Language Navigation

GridMM: Grid Memory Map for Vision-and-Language Navigation

引用

International Conference on Computer Vision (ICCV)

作者： Zihan Wang Xiangyang Li Jiahao Yang Yeqi Liu Shuqiang Jiang Key Lab of Intelligent Information Processing Laboratory of the Chinese Academy of Sciences (CAS) Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China

Vision-and-language navigation (VLN) enables the agent to navigate to a remote location following the natural language instruction in 3D environments. To represent the previously visited environment, most approaches for VLN implement memory using recurrent states, topological maps, or top-down semantic maps. In contrast to these approaches, we build the top-down egocentric and dynamically growing Grid Memory Map (i.e., GridMM) to structure the visited environment. From a global perspective, historical observations are projected into a unified grid map in a top-down view, which can better represent the spatial relations of the environment. From a local perspective, we further propose an instruction relevance aggregation method to capture fine-grained visual clues in each grid region. Extensive experiments are conducted on both the REVERIE, R2R, SOON datasets in the discrete environments, and the R2R-CE dataset in the continuous environments, showing the superiority of our proposed method. The source code is available at https://***/MrZihan/GridMM.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning from Rich Semantics and Coarse Locations for Long-tailed Object Detection

arXiv

引用

arXiv 2023年

作者： Meng, Lingchen Dai, Xiyang Yang, Jianwei Chen, Dongdong Chen, Yinpeng Liu, Mengchen Chen, Yi-Ling Wu, Zuxuan Yuan, Lu Jiang, Yu-Gang Shanghai Key Lab of Intell. Info. Processing School of CS Fudan University China Shanghai Collaborative Innovation Center of Intelligent Visual Computing China Microsoft United States

Long-tailed object detection (LTOD) aims to handle the extreme data imbalance in real-world datasets, where many tail classes have scarce instances. One popular strategy is to explore extra data with image-level labels, yet it produces limited results due to (1) semantic ambiguity—an image-level label only captures a salient part of the image, ignoring the remaining rich semantics within the image;and (2) location sensitivity—the label highly depends on the locations and crops of the original image, which may change after data transformations like random cropping. To remedy this, we propose RichSem, a simple but effective method, which is robust to learn rich semantics from coarse locations without the need of accurate bounding boxes. RichSem leverages rich semantics from images, which are then served as additional "soft supervision" for training detectors. Specifically, we add a semantic branch to our detector to learn these soft semantics and enhance feature representations for long-tailed object detection. The semantic branch is only used for training and is removed during inference. RichSem achieves consistent improvements on both overall and rare-category of LVIS under different backbones and detectors. Our method achieves state-of-the-art performance without requiring complex training and testing procedures. Moreover, we show the effectiveness of our method on other long-tailed datasets with additional experiments. Code is available at https://***/MengLcool/RichSem. Copyright © 2023, The Authors. All rights reserved.

关键词： Semantics

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：