LIVIVO - Search results -

Search results

Result 1 - 10 of total 21

Article ; Online: X

Zeng, Yan / Zhang, Xinsong / Li, Hang / Wang, Jiawei / Zhang, Jipeng / Zhou, Wangchunshu

IEEE transactions on pattern analysis and machine intelligence

2024 Volume 46, Issue 5, Page(s) 3156–3168

Abstract: Vision language pre-training aims to learn alignments between vision and language from a large amount of data. Most existing methods only learn image-text alignments. Some others utilize pre-trained object detectors to leverage vision language alignments ...

Abstract	Vision language pre-training aims to learn alignments between vision and language from a large amount of data. Most existing methods only learn image-text alignments. Some others utilize pre-trained object detectors to leverage vision language alignments at the object level. In this paper, we propose to learn multi-grained vision language alignments by a unified pre-training framework that learns multi-grained aligning and multi-grained localization simultaneously. Based on it, we present X
Language	English
Publishing date	2024-04-03
Publishing country	United States
Document type	Journal Article
ISSN	1939-3539
ISSN (online)	1939-3539
DOI	10.1109/TPAMI.2023.3339661
Database	MEDical Literature Analysis and Retrieval System OnLINE

Order via subito

This service is chargeable due to the Delivery terms set by subito. Orders including an article and supplementary material will be classified as separate orders. In these cases, fees will be demanded for each order.

Inter-library loan at ZB MED

Your chosen title can be delivered directly to ZB MED Cologne location if you are registered as a user at ZB MED Cologne.

Article ; Online: Enhancing Few-Shot Learning in Lightweight Models via Dual-Faceted Knowledge Distillation.

Zhou, Bojun / Cheng, Tianyu / Zhao, Jiahao / Yan, Chunkai / Jiang, Ling / Zhang, Xinsong / Gu, Juping

Sensors (Basel, Switzerland)

2024 Volume 24, Issue 6

Abstract: In recent computer vision research, the pursuit of improved classification performance often leads to the adoption of complex, large-scale models. However, the actual deployment of such extensive models poses significant challenges in environments ... ...

Abstract	In recent computer vision research, the pursuit of improved classification performance often leads to the adoption of complex, large-scale models. However, the actual deployment of such extensive models poses significant challenges in environments constrained by limited computing power and storage capacity. Consequently, this study is dedicated to addressing these challenges by focusing on innovative methods that enhance the classification performance of lightweight models. We propose a novel method to compress the knowledge learned by a large model into a lightweight one so that the latter can also achieve good performance in few-shot classification tasks. Specifically, we propose a dual-faceted knowledge distillation strategy that combines output-based and intermediate feature-based methods. The output-based method concentrates on distilling knowledge related to base class labels, while the intermediate feature-based approach, augmented by feature error distribution calibration, tackles the potential non-Gaussian nature of feature deviations, thereby boosting the effectiveness of knowledge transfer. Experiments conducted on MiniImageNet, CIFAR-FS, and CUB datasets demonstrate the superior performance of our method over state-of-the-art lightweight models, particularly in five-way one-shot and five-way five-shot tasks.
Language	English
Publishing date	2024-03-12
Publishing country	Switzerland
Document type	Journal Article
ZDB-ID	2052857-7
ISSN	1424-8220 ; 1424-8220
ISSN (online)	1424-8220
ISSN	1424-8220
DOI	10.3390/s24061815
Database	MEDical Literature Analysis and Retrieval System OnLINE

Order via subito

Article ; Online: Multi-Type Missing Imputation of Time-Series Power Equipment Monitoring Data Based on Moving Average Filter-Asymmetric Denoising Autoencoder.

Jiang, Ling / Gu, Juping / Zhang, Xinsong / Hua, Liang / Cai, Yueming

Sensors (Basel, Switzerland)

2023 Volume 23, Issue 24

Abstract: Supervisory control and data acquisition (SCADA) systems are widely utilized in power equipment for condition monitoring. For the collected data, there generally exists a problem-missing data of different types and patterns. This leads to the poor ... ...

Abstract	Supervisory control and data acquisition (SCADA) systems are widely utilized in power equipment for condition monitoring. For the collected data, there generally exists a problem-missing data of different types and patterns. This leads to the poor quality and utilization difficulties of the collected data. To address this problem, this paper customizes methodology that combines an asymmetric denoising autoencoder (ADAE) and moving average filter (MAF) to perform accurate missing data imputation. First, convolution and gated recurrent unit (GRU) are applied to the encoder of the ADAE, while the decoder still utilizes the fully connected layers to form an asymmetric network structure. The ADAE extracts the local periodic and temporal features from monitoring data and then decodes the features to realize the imputation of the multi-type missing. On this basis, according to the continuity of power data in the time domain, the MAF is utilized to fuse the prior knowledge of the neighborhood of missing data to secondarily optimize the imputed data. Case studies reveal that the developed method achieves greater accuracy compared to existing models. This paper adopts experiments under different scenarios to justify that the MAF-ADAE method applies to actual power equipment monitoring data imputation.
Language	English
Publishing date	2023-12-08
Publishing country	Switzerland
Document type	Journal Article
ZDB-ID	2052857-7
ISSN	1424-8220 ; 1424-8220
ISSN (online)	1424-8220
ISSN	1424-8220
DOI	10.3390/s23249697
Database	MEDical Literature Analysis and Retrieval System OnLINE

Order via subito

Article: Astronomical climate changes trigger Late Devonian bio- and environmental events in South China

Ma, Kunyuan / Hinnov, Linda / Zhang, Xinsong / Gong, Yiming

Global and planetary change. 2022 Aug., v. 215

2022

Abstract: One of the five great mass extinctions of the Phanerozoic is the Frasnian–Famennian (FF) mass extinction, for which the causes have not yet been identified. In this study, cyclostratigraphic analysis of two FF transition sections was carried out in South ...

Abstract	One of the five great mass extinctions of the Phanerozoic is the Frasnian–Famennian (FF) mass extinction, for which the causes have not yet been identified. In this study, cyclostratigraphic analysis of two FF transition sections was carried out in South China: the Yangdi section, a marine slope facies, and the Lali section, a marine basin facies. Paleoclimate proxy data collected at high resolution along these sections include magnetic susceptibility and X-ray fluorescence geochemistry (Ca and Fe concentrations). Time series analysis and modeling of the proxy data reveal that frequencies comparable to those of the Earth's long and short orbital eccentricity, obliquity, and precession index characterize the two successions. Metronomic 405-kyr long orbital eccentricity cycles identified along the two sections were used to construct a floating astronomical time scale across the FF transition, revealing that 1000 kyr separates the Lower and Upper Kellwasser horizons (LKH and UKH), and ~1600 kyr separates the maximum values of the LKH and UKH δ¹³C excursions. The estimated duration of the UKH is 150 kyr, during which the first, second, and third extinctions of the FF biotic crisis lasted 120 kyr, 20 kyr, and 10 kyr, respectively. Sedimentary noise models of the magnetic susceptibility and Ca concentration time series indicate that changes in sedimentary noise correspond to sea level variations. Modeling suggests that the long orbital eccentricity cycles controlled sea surface temperatures, and that third-order eustatic changes were forced by the combined orbital eccentricity and obliquity variations. Finally, we propose an “astronomical climate change” model as a defining mechanism of the FF biotic crisis.
Keywords	X-radiation ; basins ; climate ; climate change ; extinction ; fluorescence ; geochemistry ; magnetic susceptibility ; paleoclimatology ; sea level ; time series analysis ; China
Language	English
Dates of publication	2022-08
Publishing place	Elsevier B.V.
Document type	Article
ZDB-ID	2016967-X
ISSN	0921-8181
ISSN	0921-8181
DOI	10.1016/j.gloplacha.2022.103874
Database	NAL-Catalogue (AGRICOLA)

Order via subito

Book ; Online: Toward Building General Foundation Models for Language, Vision, and Vision-Language Understanding Tasks

Zhang, Xinsong / Zeng, Yan / Zhang, Jipeng / Li, Hang

2023

Abstract: Foundation models or pre-trained models have substantially improved the performance of various language, vision, and vision-language understanding tasks. However, existing foundation models can only perform the best in one type of tasks, namely language, ...

Abstract	Foundation models or pre-trained models have substantially improved the performance of various language, vision, and vision-language understanding tasks. However, existing foundation models can only perform the best in one type of tasks, namely language, vision, or vision-language. It is still an open question whether it is possible to construct a foundation model performing the best for all the understanding tasks, which we call a general foundation model. In this paper, we propose a new general foundation model, X-FM (the X-Foundation Model). X-FM has one language encoder, one vision encoder, and one fusion encoder, as well as a new training method. The training method includes two new techniques for learning X-FM from text, image, and image-text pair data. One is to stop gradients from the vision-language training when learning the language encoder. The other is to leverage the vision-language training to guide the learning of the vision encoder. Extensive experiments on benchmark datasets show that X-FM can significantly outperform existing general foundation models and perform better than or comparable to existing foundation models specifically for language, vision, or vision-language understanding. Code and pre-trained models are released at https://github.com/zhangxinsong-nlp/XFM.
Keywords	Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Artificial Intelligence
Subject code	004
Publishing date	2023-01-12
Publishing country	us
Document type	Book ; Online
Database	BASE - Bielefeld Academic Search Engine (life sciences selection)

Full text online

Full text

Inter-library loan at ZB MED

Your chosen title can be delivered directly to ZB MED Cologne location if you are registered as a user at ZB MED Cologne.

Article ; Online: Cross-Linking Methods of the Silk Protein Hydrogel in Oral and Craniomaxillofacial Tissue Regeneration.

Li, Xiujuan / Li, Yuanjiao / Zhang, Xinsong / Xu, Jie / Kang, Jie / Li, Bing / Zhao, Bin / Wang, Lu

Tissue engineering and regenerative medicine

2024 Volume 21, Issue 4, Page(s) 529–544

Abstract: Background: Craniomaxillofacial tissue defects are clinical defects involving craniomaxillofacial and oral soft and hard tissues. They are characterized by defect-shaped irregularities, bacterial and inflammatory environments, and the need for ... ...

Abstract	Background: Craniomaxillofacial tissue defects are clinical defects involving craniomaxillofacial and oral soft and hard tissues. They are characterized by defect-shaped irregularities, bacterial and inflammatory environments, and the need for functional recovery. Conventional clinical treatments are currently unable to achieve regeneration of high-quality oral craniomaxillofacial tissue. As a natural biomaterial, silk fibroin (SF) has been widely studied in biomedicine and has broad prospects for use in tissue regeneration. Hydrogels made of SF showed excellent water retention, biocompatibility, safety and the ability to combine with other materials. Methods: To gain an in-depth understanding of the current development of SF, this article reviews the structure, preparation and application prospects in oral and craniomaxillofacial tissue regenerative medicine. It first briefly introduces the structure of SF and then summarizes the principles, advantages and disadvantages of the different cross-linking methods (physical cross-linking, chemical cross-linking and double network structure) of SF. Finally, the existing research on the use of SF in tissue engineering and the prospects of using SF with different cross-linking methods in oral and craniomaxillofacial tissue regeneration are also discussed. Conclusions: This review is intended to show the advantages of SF hydrogels in tissue engineering and provides theoretical support for establishing novel and viable silk protein hydrogels for regeneration.
MeSH term(s)	Hydrogels/chemistry ; Humans ; Fibroins/chemistry ; Tissue Engineering/methods ; Animals ; Regeneration ; Cross-Linking Reagents/chemistry ; Regenerative Medicine/methods ; Biocompatible Materials/chemistry ; Silk/chemistry ; Mouth
Chemical Substances	Hydrogels ; Fibroins (9007-76-5) ; Cross-Linking Reagents ; Biocompatible Materials ; Silk
Language	English
Publishing date	2024-01-31
Publishing country	Korea (South)
Document type	Journal Article ; Review ; Research Support, Non-U.S. Gov't
ZDB-ID	2677535-9
ISSN	2212-5469 ; 1738-2696
ISSN (online)	2212-5469
ISSN	1738-2696
DOI	10.1007/s13770-023-00624-y
Database	MEDical Literature Analysis and Retrieval System OnLINE

Order via subito

Book ; Online: Write and Paint

Diao, Shizhe / Zhou, Wangchunshu / Zhang, Xinsong / Wang, Jiawei

Generative Vision-Language Models are Unified Modal Learners

2022

Abstract: Recent advances in vision-language pre-training have pushed the state-of-the-art on various vision-language tasks, making machines more capable of multi-modal writing (image-to-text generation) and painting (text-to-image generation). However, few ... ...

Abstract	Recent advances in vision-language pre-training have pushed the state-of-the-art on various vision-language tasks, making machines more capable of multi-modal writing (image-to-text generation) and painting (text-to-image generation). However, few studies investigate if these two essential capabilities can be learned together and boost each other, making a versatile and powerful multi-modal foundation model. In this work, we disclose the potential of symmetric generative vision-language pre-training in learning to write and paint concurrently, and propose a new unified modal model, named DaVinci, trained with prefix language modeling and prefix image modeling, a simple generative self-supervised objective on image-text pairs. Thanks to the proposed prefix multi-modal modeling framework, DaVinci is simple to train, scalable to huge data, adaptable to both writing and painting tasks, and also strong on other vision, text, and multi-modal understanding tasks. DaVinci achieves competitive performance on a wide range of 27 generation/understanding tasks and demonstrates the superiority of combining vision/language generative pre-training. Furthermore, we carefully benchmark the performance of different vision-language pre-training objectives on different scales of pre-training datasets on a heterogeneous and broad distribution coverage. Our results demonstrate the potential of exploiting self-supervision in both language and vision inputs, and establish new, stronger baselines for future comparisons at different data scales. The code and pre-trained models are available at https://github.com/shizhediao/DaVinci. Comment: ICLR 2023
Keywords	Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Computation and Language ; Computer Science - Machine Learning
Subject code	004
Publishing date	2022-06-15
Publishing country	us
Document type	Book ; Online
Database	BASE - Bielefeld Academic Search Engine (life sciences selection)

Full text online

Full text

Inter-library loan at ZB MED

Your chosen title can be delivered directly to ZB MED Cologne location if you are registered as a user at ZB MED Cologne.

Article ; Online: Robust comprehensive PV hosting capacity assessment model for active distribution networks with spatiotemporal correlation

Wu, Han / Yuan, Yue / Zhang, Xinsong / Miao, Ankang / Zhu, Junpeng

Applied Energy. 2022 Oct., v. 323 p.119558-

2022

Abstract: Variate but similar microclimate in different buses of a distribution system usually leads to correlated photovoltaic (PV) outputs. Such correlation may reduce the PV output uncertainty and enhance the hosting capacity of active distribution networks ( ... ...

Abstract	Variate but similar microclimate in different buses of a distribution system usually leads to correlated photovoltaic (PV) outputs. Such correlation may reduce the PV output uncertainty and enhance the hosting capacity of active distribution networks (ADNs). To fully elucidate the hosting capacity of geographically dispersed PV, this paper proposes a novel robust comprehensive PV capacity assessment model that considers both the spatiotemporal correlation of PV output and active distribution network management (ADNM) techniques. In the proposed model, the historical PV output data of the vicinity region are employed to generate the empirical spatial and temporal correlation matrix and ellipsoidal uncertainty sets for arbitrary PV site pairs. The uncertainty of PV outputs is addressed by a two-stage robust optimization. Concerning distribution network characteristics, the proposed capacity assessment model employs a convex conic quadratic format of AC power flow equations that are transformed into second-order cone programming. A historical PV output dataset from Suzhou China and a 59-bus rural distribution system in an adjacent city was used to demonstrate the effectiveness of the proposed PV hosting capacity assessment methodology. The hosting capacity results indicate that both spatial and temporal correlations can enhance the PV hosting capacity. Considering both the spatial and temporal leads to a significant increase in PV hosting capacity. To obtain an accurate PV hosting capacity, both spatial and temporal features should be simultaneously considered.
Keywords	data collection ; energy ; microclimate ; models ; uncertainty ; China ; Ellipsoidal uncertainty set ; PV hosting capacity ; Robust optimization ; Spatial correlation
Language	English
Dates of publication	2022-10
Publishing place	Elsevier Ltd
Document type	Article ; Online
ZDB-ID	2000772-3
ISSN	0306-2619
ISSN	0306-2619
DOI	10.1016/j.apenergy.2022.119558
Database	NAL-Catalogue (AGRICOLA)

Order via subito

Book ; Online: VLUE

Zhou, Wangchunshu / Zeng, Yan / Diao, Shizhe / Zhang, Xinsong

A Multi-Task Benchmark for Evaluating Vision-Language Models

2022

Abstract: Recent advances in vision-language pre-training (VLP) have demonstrated impressive performance in a range of vision-language (VL) tasks. However, there exist several challenges for measuring the community's progress in building general multi-modal ... ...

Abstract	Recent advances in vision-language pre-training (VLP) have demonstrated impressive performance in a range of vision-language (VL) tasks. However, there exist several challenges for measuring the community's progress in building general multi-modal intelligence. First, most of the downstream VL datasets are annotated using raw images that are already seen during pre-training, which may result in an overestimation of current VLP models' generalization ability. Second, recent VLP work mainly focuses on absolute performance but overlooks the efficiency-performance trade-off, which is also an important indicator for measuring progress. To this end, we introduce the Vision-Language Understanding Evaluation (VLUE) benchmark, a multi-task multi-dimension benchmark for evaluating the generalization capabilities and the efficiency-performance trade-off (``Pareto SOTA'') of VLP models. We demonstrate that there is a sizable generalization gap for all VLP models when testing on out-of-distribution test sets annotated on images from a more diverse distribution that spreads across cultures. Moreover, we find that measuring the efficiency-performance trade-off of VLP models leads to complementary insights for several design choices of VLP. We release the VLUE benchmark to promote research on building vision-language models that generalize well to more diverse images and concepts unseen during pre-training, and are practical in terms of efficiency-performance trade-off. Comment: ICML 2022, Benchmark website at https://vlue-benchmark.github.io
Keywords	Computer Science - Computer Vision and Pattern Recognition ; Computer Science - Computation and Language ; Computer Science - Machine Learning
Subject code	006
Publishing date	2022-05-30
Publishing country	us
Document type	Book ; Online
Database	BASE - Bielefeld Academic Search Engine (life sciences selection)

Full text online

Full text

Inter-library loan at ZB MED

Your chosen title can be delivered directly to ZB MED Cologne location if you are registered as a user at ZB MED Cologne.

Book ; Online: AMBERT

Zhang, Xinsong / Li, Hang

A Pre-trained Language Model with Multi-Grained Tokenization

2020

Abstract: Pre-trained language models such as BERT have exhibited remarkable performances in many tasks in natural language understanding (NLU). The tokens in the models are usually fine-grained in the sense that for languages like English they are words or sub- ... ...

Abstract	Pre-trained language models such as BERT have exhibited remarkable performances in many tasks in natural language understanding (NLU). The tokens in the models are usually fine-grained in the sense that for languages like English they are words or sub-words and for languages like Chinese they are characters. In English, for example, there are multi-word expressions which form natural lexical units and thus the use of coarse-grained tokenization also appears to be reasonable. In fact, both fine-grained and coarse-grained tokenizations have advantages and disadvantages for learning of pre-trained language models. In this paper, we propose a novel pre-trained language model, referred to as AMBERT (A Multi-grained BERT), on the basis of both fine-grained and coarse-grained tokenizations. For English, AMBERT takes both the sequence of words (fine-grained tokens) and the sequence of phrases (coarse-grained tokens) as input after tokenization, employs one encoder for processing the sequence of words and the other encoder for processing the sequence of the phrases, utilizes shared parameters between the two encoders, and finally creates a sequence of contextualized representations of the words and a sequence of contextualized representations of the phrases. Experiments have been conducted on benchmark datasets for Chinese and English, including CLUE, GLUE, SQuAD and RACE. The results show that AMBERT outperforms the existing best performing models in almost all cases, particularly the improvements are significant for Chinese.
Keywords	Computer Science - Computation and Language ; Computer Science - Machine Learning
Subject code	410
Publishing date	2020-08-26
Publishing country	us
Document type	Book ; Online
Database	BASE - Bielefeld Academic Search Engine (life sciences selection)

Full text online

Full text

Inter-library loan at ZB MED

Your chosen title can be delivered directly to ZB MED Cologne location if you are registered as a user at ZB MED Cologne.

To top

More links

Kategorien

Order via subito

Inter-library loan at ZB MED

More links

Kategorien

Order via subito

More links

Kategorien

Order via subito

More links

Kategorien

Order via subito

Full text online

More links

Kategorien

Inter-library loan at ZB MED

More links

Kategorien

Order via subito

Full text online

More links

Kategorien

Inter-library loan at ZB MED

More links

Kategorien

Order via subito

Full text online

More links

Kategorien

Inter-library loan at ZB MED

Full text online

More links

Kategorien

Inter-library loan at ZB MED