检索结果-内蒙古大学图书馆

International Conference on Acoustics, Speech, and Signal processing (ICASSP)

作者： Jingqi Li Jiaqi Gao Yuzhen Zhang Hongming Shan Junping Zhang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Institute of Science and Technology for Brain-inspired Intelligence Fudan University Shanghai China

As a unique biometric that can be perceived at a distance, gait has broad applications in person authentication, social security and so on. Existing gait recognition methods suffer from changes in viewpoint and clothing and barely consider extracting diverse motion features, a fundamental characteristic in gaits, from gait sequences. This paper proposes a novel motion modeling method to extract the discriminative and robust representation. Specifically, we first extract the motion features from the encoded motion sequences in the shallow layer. Then we continuously enhance the motion feature in deep layers. This motion modeling approach is independent of mainstream work in building network architectures. As a result, one can apply this motion modeling method to any backbone to improve gait recognition performance. In this paper, we combine motion modeling with one commonly used backbone (GaitGL) as GaitGL-M to illustrate motion modeling. Extensive experimental results on two commonly-used crossview gait datasets demonstrate the superior performance of GaitGL-M over existing state-of-the-art methods.

关键词： Representation learning Clothing Buildings Signal processing Network architecture Feature extraction Robustness

来源：评论

学校读者我要写书评

暂无评论

Cross-Head Supervision for Crowd Counting with Noisy Annotations

Cross-Head Supervision for Crowd Counting with Noisy Annotat...

引用

International Conference on Acoustics, Speech, and Signal processing (ICASSP)

作者： Mingliang Dai Zhizhong Huang Jiaqi Gao Hongming Shan Junping Zhang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Institute of Science and Technology for Brain-inspired Intelligence Fudan University Shanghai China

Noisy annotations such as missing annotations and location shifts often exist in crowd counting datasets due to multi-scale head sizes, high occlusion, etc. These noisy annotations severely affect the model training, especially for density map-based methods. To alleviate the negative impact of noisy annotations, we propose a novel crowd counting model with one convolution head and one transformer head, in which these two heads can supervise each other in noisy areas, called Cross-Head Supervision. The resultant model, CHS-Net, can synergize different types of inductive biases for better counting. In addition, we develop a progressive cross-head supervision learning strategy to stabilize the training process and provide more reliable supervision. Extensive experimental results on Shang-haiTech and QNRF datasets demonstrate superior performance over state-of-the-art methods. Code is available at https://***/RaccoonDML/CHSNet.

关键词： Training Adaptation models Codes Annotations Convolution Transformers Noise robustness

来源：评论

学校读者我要写书评

暂无评论

Exploring the tidal effect of urban business district with large-scale human mobility data

引用

Frontiers of computer science 2023年第3期17卷 77-90页

作者： Hongting NIU Ying SUN Hengshu ZHU Cong GENG Jiuchun YANG Hui XIONG Bo LANG School of Computer Science and Engineering Beihang UniversityBeijing 100191China Key Lab of Intelligent Information Processing of Chinese Academy of Sciences(CAS) Institute of Computing TechnologyCASBeijing 100086China Baidu Talent Intelligence Center Baidu Inc.Beijing 100085China Business School Imperial College LondonLondon SW72AZUK Artificial Intelligence Thrust The Hong Kong University of Science and TechnologyGuangzhou 510030China

Business districts are urban areas that have various functions for gathering people,such as work,consumption,leisure and *** to the dynamic nature of business activities,there exists significant tidal effect on the boundary and functionality of business ***,effectively analyzing the tidal patterns of business districts can benefit the economic and social development of a ***,with the implicit and complex nature of business district evolution,it is non-trivial for existing works to support the fine-grained and timely analysis on the tidal effect of business *** this end,we propose a data-driven and multi-dimensional framework for dynamic business district ***,we use the large-scale human trajectory data in urban areas to dynamically detect and forecast the boundary changes of business districts in different time ***,we detect and forecast the functional changes in business *** results on real-world trajectory data clearly demonstrate the effectiveness of our framework on detecting and predicting the boundary and functionality change of business ***,the analysis on practical business districts shows that our method can discover meaningful patterns and provide interesting insights into the dynamics of business *** example,the major functions of business districts will significantly change in different time periods in a day and the rate and magnitude of boundaries varies with the functional distribution of business districts.

关键词： business district trajectory functionality detection tidal effect boundary detection visiting score

来源：评论

学校读者我要写书评

暂无评论

SPEECHTOKENIZER: UNIFIED SPEECH TOKENIZER FOR SPEECH LANGUAGE MODELS

arXiv

引用

arXiv 2023年

作者： Zhang, Xin Zhang, Dong Li, Shimin Zhou, Yaqian Qiu, Xipeng School of Computer Science Fudan University Shanghai Key Laboratory of Intelligent Information Processing Fudan University

Current speech large language models build upon discrete speech representations, which can be categorized into semantic tokens and acoustic tokens. However, existing speech tokens are not specifically designed for speech language modeling. To assess the suitability of speech tokens for building speech language models, we established the first benchmark, SLMTokBench. Our results indicate that neither semantic nor acoustic tokens are ideal for this purpose. Therefore, we propose SpeechTokenizer, a unified speech tokenizer for speech large language models. SpeechTokenizer adopts the Encoder-Decoder architecture with residual vector quantization (RVQ). Unifying semantic and acoustic tokens, SpeechTokenizer disentangles different aspects of speech information hierarchically across different RVQ layers. Furthermore, We construct a Unified Speech Language Model (USLM) leveraging SpeechTokenizer. Experiments show that SpeechTokenizer performs comparably to EnCodec in speech reconstruction and demonstrates strong performance on the SLMTokBench benchmark. Also, USLM outperforms VALL-E in zero-shot Text-to-Speech tasks. Code and models are available at https://***/ZhangXInFD/SpeechTokenizer/. © 2023, CC BY-SA.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

DO-FAM: Disentangled Non-Linear Latent Navigation For Facial Attribute Manipulation

DO-FAM: Disentangled Non-Linear Latent Navigation For Facial...

引用

International Conference on Acoustics, Speech, and Signal processing (ICASSP)

作者： Yifan Yuan Siteng Ma Hongming Shan Junping Zhang School of Computer Science Shanghai Key Lab of Intelligent Information Processing Institute of Science and Technology for Brain-Inspired Intelligence Fudan University Shanghai China

Facial attribute manipulation (FAM) aims to edit the semantic attributes of facial images according to the user’s requirements. Unfortunately, the majority of existing FAM methods struggle in meeting at least one of the two requirements: high reconstruction quality and high irrelevance preservation. To alleviate these two limitations, we propose a novel Disentangled nOn-linear latent navigation framework for FAM, termed DO-FAM. To promote the reconstruction quality, we leverage hypernetworks to fine-tune a pre-trained StyleGAN2 generator. To decouple entangled attributes, we propose a novel Disentangled nOn-Linear Latent transformation module, named DOLL, which consists of three components: (1) a decomposer to factorize input latent codes into two parts: attribute-related and attribute-unrelated; (2) a non-linear Latent Transformation Network (LTNet) to navigate the attribute-related latent codes to the target one with respect to the designed attribute(s); and (3) a latent classifier tasked with predicting latent codes’ attributes to guide the latent code navigation. Extensive experimental results on a widely-used benchmark facial editing dataset, CelebA-HQ, demonstrate the superiority of our method over state-of-the-art methods.

关键词： Image quality Codes Navigation Semantics Signal processing Generators Task analysis

来源：评论

学校读者我要写书评

暂无评论

Calibrating the Confidence of Large Language Models by Eliciting Fidelity

arXiv

引用

arXiv 2024年

作者： Zhang, Mozhi Huang, Mianqiu Shi, Rundong Guo, Linsen Peng, Chong Yan, Peng Zhou, Yaqian Qiu, Xipeng School of Computer Science Fudan University China Shanghai Key Laboratory of Intelligent Information Processing Fudan University China Meituan China

Large language models optimized with techniques like RLHF have achieved good alignment in being helpful and harmless. However, post-alignment, these language models often exhibit overconfidence, where the expressed confidence does not accurately calibrate with their correctness rate. In this paper, we decompose the language model confidence into the Uncertainty about the question and the Fidelity to the answer generated by language models. Then, we propose a plug-and-play method, UF Calibration, to estimate the confidence of language models. Our method has shown good calibration performance by conducting experiments with 6 RLHF-LMs on four MCQA datasets. Moreover, we propose two novel metrics, IPR and CE, to evaluate the calibration of the model, and we have conducted a detailed discussion on Truly Well-Calibrated Confidence for large language models. Our method could serve as a strong baseline, and we hope that this work will provide some insights into the model confidence calibration. © 2024, CC BY.

关键词： Calibration

来源：评论

学校读者我要写书评

暂无评论

Temporal Feature Aggregation for Efficient 2D Video Grounding

Temporal Feature Aggregation for Efficient 2D Video Groundin...

引用

IEEE International Conference on Multimedia and Expo (ICME)

作者： Mohan Chen Yiren Zhang Jueqi Wei Yuejie Zhang Rui Feng Tao Zhang Shang Gao School of Computer Science Shanghai Key Laboratory of Intelligent Information Processing Shanghai Collaborative Innovation Center of Intelligent Visual Computing Fudan University School of Information Management and Engineering Shanghai Key Laboratory of Financial Information Technology Shanghai University of Finance and Economics School of Information Technology Deakin University

ISBN: (数字)9798350390155

ISBN: (纸本)9798350390162

Video grounding aims to locate the target video moment in an untrimmed video based on a text query. Most existing methods employ 3D CNNs as the video feature extractor, incurring substantial computational costs. Only a few methods use 2D backbones for video feature extraction, and they suffer from diminished accuracy due to the inherent lack of temporal information within 2D features. To address this problem, we propose a novel 2D video grounding method called TFA that improves accuracy while minimizing computational costs. Our approach involves a query-guided temporal feature aggregation module designed to explicitly capture temporal information. We disentangle time intervals of input video frames and prediction spans to reduce computational overhead. Additionally, we introduce deformable attention into the multi-modal encoder for further enhancement. Extensive experiments on two public datasets demonstrate that our method outperforms previous 2D video grounding methods and achieves competitive results with most 3D methods at significantly reduced costs.

关键词： Three-dimensional displays Accuracy Costs Grounding Streaming media Feature extraction Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Memory-Augmented Transformer for Efficient End-to-End Video Grounding

Memory-Augmented Transformer for Efficient End-to-End Video ...

引用

IEEE International Conference on Multimedia and Expo (ICME)

作者： Yuanwu Xu Mohan Chen Yuejie Zhang Rui Feng Tao Zhang Shang Gao School of Computer Science Shanghai Key Laboratory of Intelligent Information Processing Shanghai Collaborative Innovation Center of Intelligent Visual Computing Fudan University School of Information Management and Engineering Shanghai Key Laboratory of Financial Information Technology Shanghai University of Finance and Economics School of Information Technology Deakin University

ISBN: (数字)9798350390155

ISBN: (纸本)9798350390162

Video grounding aims to localize a specific segment corresponding to a text query in an untrimmed video. Due to the tremendous computational cost required to process the video frames, the de facto paradigm of video grounding is to extract video features using pretrained video encoders. The parameters of the video encoders are fixed during training, which limits the performance of the localization model. To solve this problem, we propose a Memory-Augmented Transformer (MAT) model. Specifically, each video is split into non-overlapping clips, and our MAT processes videos in a clip-by-clip manner while caching video features into FIFO cached memory queues. By enabling early return, our MAT outperforms previous methods with only less than 60% frames seen. Extensive experimental results on three public benchmark datasets demonstrate that our MAT can achieve competitive performance while being much more efficient than currently prevailing two-stage methods. Code is available at https://***/xuyw1997/MAT.

关键词： Training Location awareness Codes Grounding Computational modeling Streaming media Predictive models

来源：评论

学校读者我要写书评

暂无评论

Fine-Grained Blind Face Inpainting with 3D Face Component Disentanglement 48

Fine-Grained Blind Face Inpainting with 3D Face Component Di...

引用

48th IEEE International Conference on Acoustics, Speech and Signal processing, ICASSP 2023

作者： Bai, Yu He, Ruian Tan, Weimin Yan, Bo Lin, Yangle Fudan University School of Computer Science Shanghai Key Lab. of Intelligent Info. Proc. Shanghai Collab. Innov. Ctr. of Intelligent Vis. Comp. Shanghai China

ISBN: (纸本)9781728163277

Inpainting is a task to restore occlusion or other corruption on images. However, previous works require mask of the occluded area to restore the occluded image, which is inconvenient for application. Blind face inpainting aims to automatically restore the occluded face without position information of the corrupt region. In this paper, we propose a novel fine-grained blind face inpainting framework, combining 3D face components disentanglement with generative network. Canonical face texture and shape disentangled by unsupervised 3D face model is restored separately to get occlusion-free rendered result. Finally, the pixel-to-pixel generative module utilize the occlusion image and the coarse de-occlusion face to get refined inpainted result. We also build up a new dataset called CelebO-3D which consists of occluded face images synthesized with 3D occlusion and rendered by 3D face model. Extensive experiments show that the proposed method is effective and robust in face blind inpainting both in synthesized and real images. Extensive evaluations and comparison with previous methods also show our superior effectiveness and lightweight architecture. © 2023 IEEE.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

KCE-Unet: A novel music denoising method with KANConv ECA Unet*

KCE-Unet: A novel music denoising method with KANConv ECA Un...

引用

International Conference on Acoustics, Speech, and Signal processing (ICASSP)

作者： Shijie Zhang Yulun Wu Ganghui Ru Yi Yu Wei Li School of Computer Science and Technology Fudan University Shanghai China Graduate School of Advanced Science and Engineering Hiroshima University Hiroshima Japan School of Computer Science and Technology Shanghai Key Laboratory of Intelligent Information Processing Fudan University Shanghai China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

During concerts, people often spontaneously record memorable moments with their phones. However, these recordings are frequently accompanied by noise, such as cheering and applause, which diminishes the playback experience. In this paper, we introduce a novel task specifically designed for denoising music in concert environments, a challenge that has been largely overlooked in previous research. To support this task, we created a new concert denoising dataset that includes songs performed in various major languages at concerts, with noise segments like cheering and applause. Building on this, we propose KANConv ECA Unet (KCE-Unet), a method that combines the U-Net network, efficient channel attention (ECA), and the recently proposed KAN network to flexibly remove noise in the mid-to-high frequency range of spectrograms. Extensive experiments demonstrate that our method outperforms previous models in denoising performance and effectively restore disrupted musical structures.

关键词： Noise reduction Noise Buildings Acoustics Recording Multiple signal classification Speech processing Spectrogram

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：