检索结果-内蒙古大学图书馆

HYPER-REALIST RENDERING: A THEORETICAL FRAMEWORK

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Akleman, Ergun Kurt, Murat Akleman, Derya Bruins, Gary Deng, Sitong Subramanian, Meena Visual Computing & Computational Media School of PVFA Computer Science and Engineering College of Engineering Texas A&M University College StationTX77831 United States International Computer Institute Ege University Izmir Turkey Department of Statistics Texas A&M University College StationTX77831 United States Visualization Program Department of Architecture Texas A&M University College StationTX77831 United States Department of Visualization Texas A&M University College StationTX77831 United States

This is the first paper in a series on hyper-realist rendering. In this paper, we introduce the concept of hyper-realist rendering and present a theoretical framework to obtain hyper-realist images. We are using the term Hyper-realism as an umbrella word that captures all types of visual artifacts that can evoke an impression of reality. The hyper-realist artifacts are visual representations that are not necessarily created by following logical and physical principles and can still be perceived as representations of reality. This idea stems from the principles of representational arts, which attain visually acceptable renderings of scenes without implementing strict physical laws of optics and materials. The objective of this work is to demonstrate that it is possible to obtain visually acceptable illusions of reality by employing such artistic approaches. With representational art methods, we can even obtain an alternate illusion of reality that looks more real even when it is not real. This paper demonstrates that it is common to create illusions of reality in visual arts with examples of paintings by representational artists. We propose an approach to obtain expressive local and global illuminations to obtain these stylistic illusions with a set of well-defined and formal methods. Our formalism is based on four principles: (1) colors and textures must be represented by algebraically complete structures, (2) we should allow impossible and inconsistent shapes in scene concepts, (3) we have to decompose illumination and shading, and (4) we must build our shading functions using statistic-based methods instead of physics-based methods. We claim that these principles allow us to develop simple methods since we can avoid algebraic inconsistencies. Using such simple methods, it can be possible to significantly speed up modeling and rendering algorithms by focusing only on essential elements that can create illusions of reality. © 2024, CC BY.

关键词： Formal methods

Fast Collection of Target Tags Information for Large-Scale RFID Systems

学校读者我要写书评

暂无评论

Fast Collection of Target Tags Information for Large-Scale R...

International Conference on Wavelet Active Media Technology and Information Processing (ICWAMTIP)

作者： Chu Chu Ning Fu Rui Xu College of Computer Science Sichuan Normal University Chengdu China The Visual Computing and Virtual Reality Key Laboratory of Sichuan Province Chengdu China School of Network and Communication Engineering Chengdu Technological University Chengdu China School of Information and Communication Engineering University of Electronic Science and Technology of China Chengdu China

Radio Frequency Identification (RFID) technology is greatly supporting a variety of life and industry. Enhanced by sensors, RFID tags can monitor real-time information about attached objects and the environment, making efficient collection of tag information an important task. However, most works focus on collecting all tags but lack effective mechanisms to avoid interfering from irrelevant tags when the application only cares about a part of the tags in some cases. Besides, in more common scenarios with multiple RFID readers, existing works fail to effectively determine the surrounding tags for each reader, increasing the execution time. To conquer these defects, we propose a new solution called the efficient Vector-based target tag Information Collection (VIC) protocol to accelerate the collection rate. First, we introduce a composite vector to silence irrelevant tags and determine the local tags. During this process, the expected empty slots are skipped and the useless information transmitted by tags is reduced. Second, we use an allocation vector to collect tag information by resolving tag collision, thus improving the time efficiency. Finally, we provide theoretical analyses to optimize the parameter settings. Extensive experiments prove that VIC outperforms existing solutions by up to 32% increase in time efficiency.

关键词：

∂H: Differentiable Holography

学校读者我要写书评

暂无评论

Research Square

Research Square 2023年

作者： Chen, Ni Wang, Congli Heidrich, Wolfgang Wyant College of Optical Sciences University of Arizona TucsonAZ85721 United States Department of Electrical Engineering & Computer Sciences University of California BerkeleyCA94720 United States Visual Computing Center King Abdullah University of Science and Technology Thuwal23955 Saudi Arabia

Over the past decade, the field of holography has gained significant ground due to advances in computational imaging. However, the utilization of computational tools is hampered by the mismatch between experimental setups and the conceptual model. We present differentiable holography (∂H), a novel framework for automatically self-calibrating experimental imperfections in inverse holographic imaging. The technique is demonstrated on auto-focused complex field imaging from a single intensity-only inline hologram. © 2023, CC BY.

关键词： Inverse problems

Slimmable transformer with hybrid axial-attention for medical image segmentation

学校读者我要写书评

暂无评论

computers in Biology and Medicine 2024年 173卷 108370-108370页

作者： Hu, Yiyue Mu, Nan Liu, Lei Zhang, Lei Jiang, Jingfeng Li, Xiaoning College of Computer Science Sichuan Normal University Chengdu610101 China School of Science and Engineering The Chinese University of Hong Kong Shenzhen Shenzhen518172 China Department of Biomedical Engineering Michigan Technological University HoughtonMI49931 United States Visual Computing and Virtual Reality Key Laboratory of Sichuan Sichuan Normal University Chengdu610068 China Education Big Data Collaborative Innovation Center of Sichuan 2011 Chengdu610101 China

The transformer architecture has achieved remarkable success in medical image analysis owing to its powerful capability for capturing long-range dependencies. However, due to the lack of intrinsic inductive bias in modeling visual structural information, the transformer generally requires a large-scale pre-training schedule, limiting the clinical applications over expensive small-scale medical data. To this end, we propose a slimmable transformer to explore intrinsic inductive bias via position information for medical image segmentation. Specifically, we empirically investigate how different position encoding strategies affect the prediction quality of the region of interest (ROI) and observe that ROIs are sensitive to different position encoding strategies. Motivated by this, we present a novel Hybrid Axial-Attention (HAA) that can be equipped with pixel-level spatial structure and relative position information as inductive bias. Moreover, we introduce a gating mechanism to achieve efficient feature selection and further improve the representation quality over small-scale datasets. Experiments on LGG and COVID-19 datasets prove the superiority of our method over the baseline and previous works. Internal workflow visualization with interpretability is conducted to validate our success better;the proposed slimmable transformer has the potential to be further developed into a visual software tool for improving computer-aided lesion diagnosis and treatment planning. © 2024 Elsevier Ltd

关键词： COVID-19

A non-asymptotic analysis for stein variational gradient descent 34

学校读者我要写书评

暂无评论

A non-asymptotic analysis for stein variational gradient des...

34th Conference on Neural Information Processing Systems, NeurIPS 2020

作者： Korba, Anna Salim, Adil Arbel, Michael Luise, Giulia Gretton, Arthur Gatsby Computational Neuroscience Unit University College London United Kingdom Visual Computing Center KAUST Computer Science Department University College London United Kingdom

We study the Stein Variational Gradient Descent (SVGD) algorithm, which optimises a set of particles to approximate a target probability distribution p ? e-V on Rd. In the population limit, SVGD performs gradient descent in the space of probability distributions on the KL divergence with respect to p, where the gradient is smoothed through a kernel integral operator. In this paper, we provide a novel finite time analysis for the SVGD algorithm. We provide a descent lemma establishing that the algorithm decreases the objective at each iteration, and rates of convergence for the averaged Stein Fisher divergence (also referred to as Kernel Stein Discrepancy). We also provide a convergence result of the finite particle system corresponding to the practical implementation of SVGD to its population version. © 2020 Neural information processing systems foundation. All rights reserved.

关键词： Probability distributions

AlphaChimp: Tracking and Behavior Recognition of Chimpanzees

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Ma, Xiaoxuan Lin, Yutang Xu, Yuan Kaufhold, Stephan P. Terwilliger, Jack Meza, Andres Zhu, Yixin Rossano, Federico Wang, Yizhou Center on Frontiers of Computing Studies School of Computer Science Peking University China Yuanpei College Peking University China Department of Cognitive Science University of California San Diego United States Institute for Artificial Intelligence Peking University China Nat'l Eng. Research Center of Visual Technology Nat'l Key Lab of General Artificial Intelligence Peking University China

Understanding non-human primate behavior is crucial for improving animal welfare, modeling social behavior, and gaining insights into both distinctly human and shared behaviors. Despite recent advances in computer vision, automated analysis of primate behavior remains challenging due to the complexity of their social interactions and the lack of specialized algorithms. Existing methods often struggle with the nuanced behaviors and frequent occlusions characteristic of primate social dynamics. This study aims to develop an effective method for automated detection, tracking, and recognition of chimpanzee behaviors in video footage. Here we show that our proposed method, AlphaChimp, an end-to-end approach that simultaneously detects chimpanzee positions and estimates behavior categories from videos, significantly outperforms existing methods in behavior recognition. AlphaChimp achieves approximately 10% higher tracking accuracy and a 20% improvement in behavior recognition compared to state-of-the-art methods, particularly excelling in the recognition of social behaviors. This superior performance stems from AlphaChimp's innovative architecture, which integrates temporal feature fusion with a Transformer-based self-attention mechanism, enabling more effective capture and interpretation of complex social interactions among chimpanzees. Our approach bridges the gap between computer vision and primatology, enhancing technical capabilities and deepening our understanding of primate communication and sociality. We release our code and models at project page and hope this will facilitate future research in animal social dynamics. This work contributes to ethology, cognitive science, and artificial intelligence, offering new perspectives on social intelligence. Copyright © 2024, The Authors. All rights reserved.

关键词： Mammals

Deploying Hybrid MAELM Approach for Human Emotion Detection Through Speech and Facial Expressions

学校读者我要写书评

暂无评论

Deploying Hybrid MAELM Approach for Human Emotion Detection ...

International Conference on Innovative Mechanisms for Industry Applications (ICIMIA)

作者： K Senthil Kumar S. Rukmani Devi Nidhi Ranjan Gitika Rath G. Indira Neerav Nishant Department of Networking and Communications School of Computing College of Engineering and Technology SRM Institute of Science and Technology Chennai India Department of Computer Science Saveetha College of Liberal Arts and Sciences SIMATS Deemed to be University Chennai India Computer Engineering Vasantdada Patil Pratishthan’s College of Engineering and Visual Arts Mumbai India Department of CSE Marri Laxman Reddy Institute of Technology and Management Hyderabad India Department of Electrical and Electronics Engineering Prince Shri Venkateshwara Padmavathy Engineering College Chennai India Department of Computer Science and Engineering School of Engineering Babu Banarasi Das University Lucknow India

Emotion recognition is one of the most fascinating emerging fields of research. It’s useful in a lot of different contexts. Some of the most exciting applications of this technology involve robots’ ability to see and communicate with one another. Human emotions can be recognized through both visual and audible cues. Facial expressions are one of the most accurate indicators of a person’s emotional state. Data preprocessing, feature extraction, and model training are the first three steps of the proposed methodology. Image resizing, a median filter, noise removal, and histogram equalization are all components of preprocessing. Gaussian mixture models and the gray level co-occurrence matrix are being used for feature extraction. After the extraction of features, models with those features are trained using MA-ELM. The proposed method outperforms PCA and ELM, two of the most popular alternatives.

关键词：

Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Huang, Bingchen Chen, Zhineng Zhou, Peng Chen, Jiayin Wu, Zuxuan Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University China Shanghai Collaborative Innovation Center on Intelligent Visual Computing China University of Maryland College ParkMD United States

The dynamic expansion architecture is becoming popular in class incremental learning, mainly due to its advantages in alleviating catastrophic forgetting. However, task confusion is not well assessed within this framework, e.g., the discrepancy between classes of different tasks is not well learned (i.e., inter-task confusion, ITC), and certain priority is still given to the latest class batch (i.e., old-new confusion, ONC). We empirically validate the side effects of the two types of confusion. Meanwhile, a novel solution called Task Correlated Incremental Learning (TCIL) is proposed to encourage discriminative and fair feature utilization across tasks. TCIL performs a multi-level knowledge distillation to propagate knowledge learned from old tasks to the new one. It establishes information flow paths at both feature and logit levels, enabling the learning to be aware of old classes. Besides, attention mechanism and classifier re-scoring are applied to generate more fair classification scores. We conduct extensive experiments on CIFAR100 and ImageNet100 datasets. The results demonstrate that TCIL consistently achieves state-of-the-art accuracy. It mitigates both ITC and ONC, while showing advantages in battle with catastrophic forgetting even no rehearsal memory is reserved. 1 Copyright © 2022, The Authors. All rights reserved.

关键词： Distillation

Consistent two-flow network for tele-registration of point clouds

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Yan, Zihao Yi, Zimu Hu, Ruizhen Mitra, Niloy J. Cohen-Or, Daniel Huang, Hui Visual Computing Research Center College of Computer Science and Software Engineering Shenzhen University University College London Shenzhen University Tel Aviv University

Rigid registration of partial observations is a fundamental problem in various applied fields. In computer graphics, special attention has been given to the registration between two partial point clouds generated by scanning devices. State-of-the-art registration techniques still struggle when the overlap region between the two point clouds is small, and completely fail if there is no overlap between the scan pairs. In this paper, we present a learning-based technique that alleviates this problem, and allows registration between point clouds, presented in arbitrary poses, and having little or even no overlap, a setting that has been referred to as tele-registration. Our technique is based on a novel neural network design that learns a prior of a class of shapes and can complete a partial shape. The key idea is combining the registration and completion tasks in a way that reinforces each other. In particular, we simultaneously train the registration network and completion network using two coupled flows, one that register-and-complete, and one that complete-and-register, and encourage the two flows to produce a consistent result. We show that, compared with each separate flow, this two-flow training leads to robust and reliable tele-registration, and hence to a better point cloud prediction that completes the registered scans. It is also worth mentioning that each of the components in our neural network outperforms state-of-the-art methods in both completion and registration. We further analyze our network with several ablation studies and demonstrate its performance on a large number of partial point clouds, both synthetic and real-world, that have only small or no overlap. Copyright © 2021, The Authors. All rights reserved.

关键词： Neural networks