检索结果-内蒙古大学图书馆

15th Asian Conference on Computer Vision, ACCV 2020

作者： Guo, Zidong Yuan, Zejian Zhang, Chong Chi, Wanchao Ling, Yonggen Zhang, Shenghao Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University Xi’an China Tencent Robotics X Beijing China

ISBN: (纸本)9783030695408

Gaze is the essential manifestation of human attention. In recent years, a series of work has achieved high accuracy in gaze estimation. However, the inter-personal difference limits the reduction of the subject-independent gaze estimation error. This paper proposes an unsupervised method for domain adaptation gaze estimation to eliminate the impact of inter-personal diversity. In domain adaption, we design an embedding representation with prediction consistency to ensure that linear relationships between gaze directions in different domains remain consistent on gaze space and embedding space. Specifically, we employ source gaze to form a locally linear representation in the gaze space for each target domain prediction. Then the same linear combinations are applied in the embedding space to generate hypothesis embedding for the target domain sample, remaining prediction consistency. The deviation between the target and source domain is reduced by approximating the predicted and hypothesis embedding for the target domain sample. Guided by the proposed strategy, we design Domain Adaptation Gaze Estimation Network(DAGEN), which learns embedding with prediction consistency and achieves state-of-the-art results on both the MPIIGaze and the EYEDIAP datasets. © 2021, Springer Nature Switzerland AG.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

NeuralDiffuser: Neuroscience-inspired Diffusion Guidance for fMRI Visual Reconstruction

arXiv

引用

arXiv 2024年

作者： Li, Haoyu Wu, Hao Chen, Badong Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University China School of Electrical Engineering Xi’an University of Technology China

Reconstructing visual stimuli from functional Magnetic Resonance Imaging (fMRI) enables fine-grained retrieval of brain activity. However, the accurate reconstruction of diverse details, including structure, background, texture, color, and more, remains challenging. The stable diffusion models inevitably result in the variability of reconstructed images, even under identical conditions. To address this challenge, we first uncover the neuroscientific perspective of diffusion methods, which primarily involve top-down creation using pre-trained knowledge from extensive image datasets, but tend to lack detail-driven bottom-up perception, leading to a loss of faithful details. In this paper, we propose NeuralDiffuser, which incorporates primary visual feature guidance to provide detailed cues in the form of gradients. This extension of the bottom-up process for diffusion models achieves both semantic coherence and detail fidelity when reconstructing visual stimuli. Furthermore, we have developed a novel guidance strategy for reconstruction tasks that ensures the consistency of repeated outputs with original images rather than with various outputs. Extensive experimental results on the Natural Senses Dataset (NSD) qualitatively and quantitatively demonstrate the advancement of NeuralDiffuser by comparing it against baseline and state-of-the-art methods horizontally, as well as conducting longitudinal ablation studies. Code can be available on https://***/HaoyyLi/NeuralDiffuser. Copyright © 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

Jacobian Regularizer-based Neural Granger Causality

arXiv

引用

arXiv 2024年

作者： Zhou, Wanqi Bai, Shuanghao Yu, Shujian Zhao, Qibin Chen, Badong Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University China RIKEN AIP Japan Vrije Universiteit Amsterdam Netherlands

With the advancement of neural networks, diverse methods for neural Granger causality have emerged, which demonstrate proficiency in handling complex data, and nonlinear relationships. However, the existing framework of neural Granger causality has several limitations. It requires the construction of separate predictive models for each target variable, and the relationship depends on the sparsity on the weights of the first layer, resulting in challenges in effectively modeling complex relationships between variables as well as unsatisfied estimation accuracy of Granger causality. Moreover, most of them cannot grasp full-time Granger causality. To address these drawbacks, we propose a Jacobian Regularizer-based Neural Granger Causality (JRNGC) approach, a straightforward yet highly effective method for learning multivariate summary Granger causality and full-time Granger causality by constructing a single model for all target variables. Specifically, our method eliminates the sparsity constraints of weights by leveraging an input-output Jacobian matrix regularizer, which can be subsequently represented as the weighted causal matrix in the post-hoc analysis. Extensive experiments show that our proposed approach achieves competitive performance with the state-of-the-art methods for learning summary Granger causality and full-time Granger causality while maintaining lower model complexity and high scalability. Copyright © 2024, The Authors. All rights reserved.

关键词： Jacobian matrices

来源：评论

学校读者我要写书评

暂无评论

Design of a Water Surface Robot Photographer with the Optimized Interface Photography Algorithm*

Design of a Water Surface Robot Photographer with the Optimi...

引用

IEEE International Conference on Real-time Computing and robotics (RCAR)

作者： Cheng Liang Minen Lv Xu Dai Huihuan Qian Shenzhen Institute of Artificial Intelligence and Robotics for Society The Chinese University of Hong Kong Shenzhen Huzhou Institute of Quality and Technical Supervision and Inspection

ISBN: (数字)9798350372601

ISBN: (纸本)9798350372618

Robot photographers are becoming increasingly popular for capturing images from unique perspectives. However, there are still many technical challenges to be addressed for automatic simultaneous photography of the above and below-water parts of objects floating on the water surface. In this paper, we present a three-thruster surface robot photographer with an algorithm for processing floating objects’ photos on both sides of the water surface. We also propose a strategy for autonomous navigation and shooting while surrounding floating objects. We combine various computer image processing algorithms to solve various shooting subtasks, such as water surface position recognition, image stitching on both sides of the water surface, and local photo stitching of large boats or other floating objects. Additionally, we propose a sea-state-adaptive photography optimization algorithm to obtain qualified, stable, and clear photos of floating objects in wave environments. Our experimental results demonstrate that the proposed system can effectively capture high-quality images of floating objects from both above and below the water surface.

关键词： Photography Sea surface Image recognition Surface waves Boats Real-time systems Surface treatment

来源：评论

学校读者我要写书评

暂无评论

Automatic Generation of Frontend Code from Design Interface 13

Automatic Generation of Frontend Code from Design Interface

引用

13th International Conference on Information Science and Technology, ICIST 2023

作者： He, Xianyi Wang, Jize Liu, Zhe Xie, Le Wang, Hesheng Le, Xinyi Department of Automation Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China Institute of Medical Robotics Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9798350313925

Internet technology is constantly evolving, and application software plays a crucial role in various fields. The development of various software relies heavily on the design of user interfaces and the corresponding code implementation. However, implementing front-end code is a highly repetitive and tedious task. Prior to designing the front-end interface based on the requirements document, designers usually create a simpler interface to assist in designing and implementing the actual interface. The specific focus of this study is to generate a domain-specific language (DSL) file based on a screenshot of the frontend interface using deep learning. The pix2code project was the first to propose using deep learning for DSL text generation. However, the pix2code project has several issues: firstly, the DSL language designed in the project lacks an adequate description vocabulary to meet the basic requirements of describing actual web pages;secondly, the project has poor scalability for the DSL language;and finally, the project's generation effectiveness using deep learning models is average. This study expands on the pix2code project by extending the design of the domain-specific language, introducing new descriptive vocabulary, and developing an algorithm to generate a new dataset. These enhancements allow the DSL to better describe various components and their combinations within the interface. Additionally, three improvements are made to the original deep learning model. Firstly, the convolutional network used for image feature extraction is deepened to enhance feature extraction capabilities. Secondly, a text encoder based on Transformer is designed for text feature extraction. Finally, a bidirectional LSTM network is used for decoding. Experimental results show that the proposed model in this study improves the BLEU score from 0.81 to 0.85 on a publicly available dataset, and from 0.547 to 0.575 on a newly created dataset. © 2023 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Unbiased Scene Graph Generation Via Head-Tail Cooperative Network with Self-Supervised Learning

SSRN

引用

SSRN 2024年

作者： Wang, Lei Yuan, Zejian Lu, Yao Chen, Badong Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University Xi’an710049 China Beijing Institute of Remote Sensing Beijing100011 China

Scene Graph Generation (SGG) as a critical task in image understanding, facing the challenge of head-biased prediction caused by the long-tail distribution of predicates. However, current debiased SGG methods can easily prioritize improving the prediction of tail predicates while ignoring the substantial sacrifice of head predicates, leading to a shift from head bias to tail bias. To address this issue, we propose a Head-Tail Cooperative network with self-supervised Learning (HTCL), which achieves unbiased SGG by cooperating head-prefer and tail-prefer predictions through learnable weight parameters. HTCL employs a tail-prefer feature encoder to re-represent predicate features by injecting self-supervised learning, which focuses on the intrinsic structure of features, into the supervised learning of SGG, constraining the representation of predicate features to enhance the distinguishability of tail samples. We demonstrate the effectiveness of our HTCL by applying it to VG150, Open Images V6 and GQA200 datasets. The results show that HTCL achieves higher mean Recall with a minimal sacrifice in Recall and achieves a new state-of-the-art overall performance. Our code is available at https://***/wanglei0618/HTCL. © 2024, The Authors. All rights reserved.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

Continual Forgetting for Pre-trained Vision Models

arXiv

引用

arXiv 2024年

作者： Zhao, Hongbo Ni, Bolin Fan, Junsong Wang, Haochen Wang, Yuxi Zhu, Fei Chen, Yuntao Meng, Gaofeng Zhang, Zhaoxiang State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences China Centre for Artificial Intelligence and Robotics Hong Kong Institute of Science & Innovation Chinese Academy of Sciences China University of Chinese Academy of Sciences China Shanghai Artificial Intelligence Laboratory China

For privacy and security concerns, the need to erase unwanted information from pre-trained vision models is becoming evident nowadays. In real-world scenarios, erasure requests originate at any time from both users and model owners. These requests usually form a sequence. Therefore, under such a setting, selective information is expected to be continuously removed from a pre-trained model while maintaining the rest. We define this problem as continual forgetting and identify two key challenges. (i) For unwanted knowledge, efficient and effective deleting is crucial. (ii) For remaining knowledge, the impact brought by the forgetting procedure should be minimal. To address them, we propose Group Sparse LoRA (GS-LoRA). Specifically, towards (i), we use LoRA modules to fine-tune the FFN layers in Transformer blocks for each forgetting task independently, and towards (ii), a simple group sparse regularization is adopted, enabling automatic selection of specific LoRA groups and zeroing out the others. GS-LoRA is effective, parameter-efficient, data-efficient, and easy to implement. We conduct extensive experiments on face recognition, object detection and image classification and demonstrate that GS-LoRA manages to forget specific classes with minimal impact on other classes. Codes will be released on https://***/bjzhb666/GS-LoRA. Copyright © 2024, The Authors. All rights reserved.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Unified regularity measures for sample-wise learning and generalization

引用

Visual intelligence 2024年第1期2卷 1-20页

作者： Zhang, Chi Yuan, Meng Ma, Xiaoning Liu, Yu Lu, Haoang Wang, Le Su, Yuanqi Liu, Yuehu Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University Xi’an 710049 China National KeyLaboratory of Human-Machine Hybrid Augmented Intelligence Xi’an Jiaotong University Xi’an 710049 China National Engineering Research Center for Visual Information and Applications Xi’an Jiaotong University Xi’an 710049 China School of Computer Science and Technology Xi’an Jiaotong University Xi’an 710049 China

Fundamental machine learning theory shows that different samples contribute unequally to both the learning and testing processes. Recent studies on deep neural networks (DNNs) suggest that such sample differences are rooted in the distribution of intrinsic pattern information, namely sample regularity. Motivated by recent discoveries in network memorization and generalization, we propose a pair of sample regularity measures with a formulation-consistent representation for both processes. Specifically, the cumulative binary training/generalizing loss (CBTL/CBGL), the cumulative number of correct classifications of the training/test sample within the training phase, is proposed to quantify the stability in the memorization-generalization process, while forgetting/mal-generalizing events (ForEvents/MgEvents), i.e., the misclassification of previously learned or generalized samples, are utilized to represent the uncertainty of sample regularity with respect to optimization dynamics. The effectiveness and robustness of the proposed approaches for mini-batch stochastic gradient descent (SGD) optimization are validated through sample-wise analyses. Further training/test sample selection applications show that the proposed measures, which share the unified computing procedure, could benefit both tasks. © The Author(s) 2024.

关键词： Forgetting events Memorization-generalization mechanism Network generalization Sample regularity Sample reweighting

来源：评论

学校读者我要写书评

暂无评论

Convolutional Recurrent MetriCGAN With Spectral Dimension Compression For Full-Band Speech Enhancement 48

Convolutional Recurrent MetriCGAN With Spectral Dimension Co...

引用

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

作者： Hou, Zhongshu Hu, Qinwen Sun, Tianchi Hu, Yuxiang Zhu, Changbao Chen, Kai Nanjing University Key Laboratory of Modern Acoustics Nanjing210093 China NJU-Horizon Intelligent Audio Lab Horizon Robotics Beijing100094 China Nanjing Institute of Advanced Artificial Intelligence Nanjing210014 China

ISBN: (纸本)9781728163277

MetricGAN and its variations have been proven to be an effective wide-band speech enhancement model. In this paper, we expand it to full-band enhancement by combining our recently proposed learnable spectral dimension compression mapping strategy. The encoder-decoder structure with a time-frequency convolutional recurrent network is utilized as the generator. The proposed model is submitted to the ICASSP Signal Processing Grand Challenge: DNS-5 Challenge (2023). Without using the enrollment speech, it obtains a final score of 0.548 on Track-1 and 0.559 on Track-2. © 2023 IEEE.

关键词： CMGAN Full-band speech enhancement MetricGAN spectral dimension compression mapping

来源：评论

学校读者我要写书评

暂无评论

A Survey on NIST Selected Third Round Candidates for Post Quantum Cryptography 7

A Survey on NIST Selected Third Round Candidates for Post Qu...

引用

7th International Conference on Communication and Electronics Systems, ICCES 2022

作者： Hasija, Taniya Ramkumar, K.R. Kaur, Amanpreet Mittal, Sudesh Singh, Bhupendra Chitkara University Institute of Engineering and Technology Chitkara University Punjab India Defence Research and Development Organisation Centre for Artificial Intelligence and Robotics Bangalore India

ISBN: (数字)9781665496346

ISBN: (纸本)9781665496346

The tremendous increase in the volume of data over internet for many applications such as banking, hospitals, insurance, online trading, messengers, voice communications and others has resulted in increased security threats. Due to the evaluation in the quantum computer generation, the classical cryptography algorithm will be broken in very less amount of time and lead to the security breaches and malicious attacks. The National institute of Standard and Technology (NIST) has invited a new generation of Quantum Safe cryptography. First two rounds are completed successfully, and the finalist has been announced. This paper provides an in-depth examination of NIST third round selected candidates which are going to be the standard Post Quantum Cryptography (PQC) algorithm. Also, the limitations and attacking concerns related to that PQC algorithms are presented. © 2022 IEEE.

关键词： Quantum cryptography

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：