Search

Your search keyword '"Sun, Shikun"' showing total 131 results
Start Over Author "Sun, Shikun" Search Limiters Full Text
131 results on '"Sun, Shikun"'

1. DanceCamAnimator: Keyframe-Based Controllable 3D Dance Camera Synthesis

Author

Wang, Zixuan, Li, Jiayi, Qin, Xiaoyu, Sun, Shikun, Zhou, Songtao, Jia, Jia, and Luo, Jiebo

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Multimedia
Abstract

Synthesizing camera movements from music and dance is highly challenging due to the contradicting requirements and complexities of dance cinematography. Unlike human movements, which are always continuous, dance camera movements involve both continuous sequences of variable lengths and sudden drastic changes to simulate the switching of multiple cameras. However, in previous works, every camera frame is equally treated and this causes jittering and unavoidable smoothing in post-processing. To solve these problems, we propose to integrate animator dance cinematography knowledge by formulating this task as a three-stage process: keyframe detection, keyframe synthesis, and tween function prediction. Following this formulation, we design a novel end-to-end dance camera synthesis framework \textbf{DanceCamAnimator}, which imitates human animation procedures and shows powerful keyframe-based controllability with variable lengths. Extensive experiments on the DCM dataset demonstrate that our method surpasses previous baselines quantitatively and qualitatively. Code will be available at \url{https://github.com/Carmenw1203/DanceCamAnimator-Official}., Comment: Accepted by ACM Multimedia 2024

Published
2024
Full Text
View/download PDF

2. PlacidDreamer: Advancing Harmony in Text-to-3D Generation

Author

Huang, Shuo, Sun, Shikun, Wang, Zixuan, Qin, Xiaoyu, Xiong, Yanmin, Zhang, Yuan, Wan, Pengfei, Zhang, Di, and Jia, Jia

Subjects
Computer Science - Computer Vision and Pattern Recognition, I.4.0
Abstract

Recently, text-to-3D generation has attracted significant attention, resulting in notable performance enhancements. Previous methods utilize end-to-end 3D generation models to initialize 3D Gaussians, multi-view diffusion models to enforce multi-view consistency, and text-to-image diffusion models to refine details with score distillation algorithms. However, these methods exhibit two limitations. Firstly, they encounter conflicts in generation directions since different models aim to produce diverse 3D assets. Secondly, the issue of over-saturation in score distillation has not been thoroughly investigated and solved. To address these limitations, we propose PlacidDreamer, a text-to-3D framework that harmonizes initialization, multi-view generation, and text-conditioned generation with a single multi-view diffusion model, while simultaneously employing a novel score distillation algorithm to achieve balanced saturation. To unify the generation direction, we introduce the Latent-Plane module, a training-friendly plug-in extension that enables multi-view diffusion models to provide fast geometry reconstruction for initialization and enhanced multi-view images to personalize the text-to-image diffusion model. To address the over-saturation problem, we propose to view score distillation as a multi-objective optimization problem and introduce the Balanced Score Distillation algorithm, which offers a Pareto Optimal solution that achieves both rich details and balanced saturation. Extensive experiments validate the outstanding capabilities of our PlacidDreamer. The code is available at \url{https://github.com/HansenHuang0823/PlacidDreamer}., Comment: Accepted by ACM Multimedia 2024

Published
2024

3. DanceCamera3D: 3D Camera Movement Synthesis with Music and Dance

Author

Wang, Zixuan, Jia, Jia, Sun, Shikun, Wu, Haozhe, Han, Rong, Li, Zhenyu, Tang, Di, Zhou, Jiaqing, and Luo, Jiebo

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Multimedia
Abstract

Choreographers determine what the dances look like, while cameramen determine the final presentation of dances. Recently, various methods and datasets have showcased the feasibility of dance synthesis. However, camera movement synthesis with music and dance remains an unsolved challenging problem due to the scarcity of paired data. Thus, we present DCM, a new multi-modal 3D dataset, which for the first time combines camera movement with dance motion and music audio. This dataset encompasses 108 dance sequences (3.2 hours) of paired dance-camera-music data from the anime community, covering 4 music genres. With this dataset, we uncover that dance camera movement is multifaceted and human-centric, and possesses multiple influencing factors, making dance camera synthesis a more challenging task compared to camera or dance synthesis alone. To overcome these difficulties, we propose DanceCamera3D, a transformer-based diffusion model that incorporates a novel body attention loss and a condition separation strategy. For evaluation, we devise new metrics measuring camera movement quality, diversity, and dancer fidelity. Utilizing these metrics, we conduct extensive experiments on our DCM dataset, providing both quantitative and qualitative evidence showcasing the effectiveness of our DanceCamera3D model. Code and video demos are available at https://github.com/Carmenw1203/DanceCamera3D-Official., Comment: Accept to CVPR 2024

Published
2024

4. SDDM: Score-Decomposed Diffusion Models on Manifolds for Unpaired Image-to-Image Translation

Author

Sun, Shikun, Wei, Longhui, Xing, Junliang, Jia, Jia, and Tian, Qi

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Recent score-based diffusion models (SBDMs) show promising results in unpaired image-to-image translation (I2I). However, existing methods, either energy-based or statistically-based, provide no explicit form of the interfered intermediate generative distributions. This work presents a new score-decomposed diffusion model (SDDM) on manifolds to explicitly optimize the tangled distributions during image generation. SDDM derives manifolds to make the distributions of adjacent time steps separable and decompose the score function or energy guidance into an image ``denoising" part and a content ``refinement" part. To refine the image in the same noise level, we equalize the refinement parts of the score function and energy guidance, which permits multi-objective optimization on the manifold. We also leverage the block adaptive instance normalization module to construct manifolds with lower dimensions but still concentrated with the perturbed reference image. SDDM outperforms existing SBDM-based methods with much fewer diffusion steps on several I2I benchmarks.

Published
2023

5. ERNIE-Search: Bridging Cross-Encoder with Dual-Encoder via Self On-the-fly Distillation for Dense Passage Retrieval

Author

Lu, Yuxiang, Liu, Yiding, Liu, Jiaxiang, Shi, Yunsheng, Huang, Zhengjie, Sun, Shikun Feng Yu, Tian, Hao, Wu, Hua, Wang, Shuaiqiang, Yin, Dawei, and Wang, Haifeng

Subjects
Computer Science - Computation and Language
Abstract

Neural retrievers based on pre-trained language models (PLMs), such as dual-encoders, have achieved promising performance on the task of open-domain question answering (QA). Their effectiveness can further reach new state-of-the-arts by incorporating cross-architecture knowledge distillation. However, most of the existing studies just directly apply conventional distillation methods. They fail to consider the particular situation where the teacher and student have different structures. In this paper, we propose a novel distillation method that significantly advances cross-architecture distillation for dual-encoders. Our method 1) introduces a self on-the-fly distillation method that can effectively distill late interaction (i.e., ColBERT) to vanilla dual-encoder, and 2) incorporates a cascade distillation process to further improve the performance with a cross-encoder teacher. Extensive experiments are conducted to validate that our proposed solution outperforms strong baselines and establish a new state-of-the-art on open-domain QA benchmarks.

Published
2022

18. Inter‐Regional Food‐Water‐Income Synergy Through Bi‐Level Crop Redistribution Model Coupled With Virtual Water: A Case Study of China's Hetao Irrigation District.

Author

Yin, Jieling, Li, Xin, Engel, Bernie A., Ding, Jiayi, Xing, Xin, Sun, Shikun, and Wang, Yubao

Subjects
WATER management, WATER efficiency, ENVIRONMENTAL security, WATER use, IRRIGATION, WATER security
Abstract

Incorporating water footprints and virtual water into crop redistribution provides a new approach for efficient water resources utilization and synergistic development of water surplus and scarce regions. In this work, the absolute and comparative advantage of the production‐based blue and gray water footprint (PWFblue and PWFgray), the calorie‐based blue water footprint (CWFblue) and the net benefit‐based blue water footprint (NBWFblue) were used as coefficients to establish a bi‐level crop redistribution model. The mode considers upper‐level decision makers interested in maximizing food security and ecological security and lower‐level decision makers interested in water use efficiency, water use benefits and net benefits. The model was applied in the Hetao Irrigation District (HID), China. The results showed that after optimization, the PWFblue, CWFblue, NBWFblue, and gray water footprint (GWF) of the HID were reduced by 23.32%, 5.60%, 17.40%, and 6.67%, respectively. National benefits were improved, especially when considering synergistic optimization, although the net benefits of HID was affected. The calorie supply increased by 9.6 × 109 kcal, the GWF decreased by 8.29 × 106 m3, and water use efficiency and benefits were improved in China. In contrast, the calorie supply and the net benefits of the HID decreased, while the GWF increased. Moreover, multiple stakeholders were involved in crop redistribution and required national synergies. The bi‐level model proved more suitable than the multi‐objective model. The model proposed in this work considers synergies outside the region in crop redistribution within the region, and can provide new insight for water and soil resources management in arid and semi‐arid regions. Key Points: Virtual water flow embedded in optimization model reflecting comparative advantageAbsolute advantage and comparative advantage synergize interregional interestsBi‐level optimization model trade‐offs regional authority and sub‐regions [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

34. Continuous film mulching enhances maize root growth and grain yield by improving soil aggregates characteristic in semiarid areas of China: An eight-year field experiment

Author

yang jialin, Zhou Yanduo, Ye Xu, sun shikun, liu enke, ren xiaolong, Jia zhikuan, Wei ting, and Peng Zhang

Abstract

AimsFilm mulching has been widely used in agricultural production in semiarid areas. While studies regarding the long-term effects of different film mulching methods on soil aggregates and stability are still lacking. MethodsIn the Loess Plateau of China, a 8-year continuous field experiment was conducted on soil aggregate and stability, and crop yield with five different film mulching treatments: (i) a control method, flat planting without mulching (CK); (ii) alternate mulching and bare rows without ridges (P); (iii) planting furrows separated by consecutive plastic film-mulched ridges (S); (iv) double ridge-furrow with full plastic film mulching (D); and (v) traditional ridge-furrow rainwater harvesting planting (R). The indexes were used to expressed soil aggregate stability by mean-weight diameter (MWD), geometric mean diameter (GMD), fractal dimension (FD), percentage of aggregate destruction (PAD) and unstable aggregates index (ELT).ResultsAfter 8 years continuous film mulching, soil quality was significantly improved. Compared with CK, mulching could decrease soil bulk density (2.96%-4.54%), and increase soil porosity (2.45%-4.00%), soil macroaggregate content (7.39%-22.92% by dry-sieving, 8.49%-21.67% by wet-sieving), soil aggregate stability and crop yield (29.98%-63.96%), especially under S and D treatments. D and S significantly improved soil structure (soil aggregate content and stability) with the increase of crop root biomass (43.59%, and 47.25%, respectively), compared with other mulching treatments. The root system characteristic parameters under S and D treatment were significantly higher than R and P, especially during 60–120 days after planting.ConclusionsOn the whole, the S treatment has a greater effect on crop yield increase and soil quality improve, and thus it was suggested as an effective mulching pattern on stimulate and stable crop yields in the Loess Plateau of China.

Published
2022

38. Characterisation of Staphylococcus aureus strain causing severe respiratory disease in rabbits

Author

Sang Lei, Chen Yanfeng, Chen Dongjin, Sun Shikun, Wang Jinxiang, and Xie Xiping

Subjects
Veterinary medicine, Staphylococcus aureus, Meticillin, rabbit, virulence factors, Rabbit, medicine.disease_cause, 03 medical and health sciences, medicine, multi-locus sequencing typing, Respiratory disease, 030304 developmental biology, lcsh:SF1-1100, 0303 health sciences, Antiinfective agent, lcsh:Veterinary medicine, Virulence factors, 030306 microbiology, business.industry, Drug susceptibility, medicine.disease, Pathogenicity, respiratory disease, Technology system, lcsh:SF600-1100, Animal Science and Zoology, lcsh:Animal culture, business, Multi-locus sequencing typing, medicine.drug, Beta lactam antibiotics
Abstract

[EN] Staphylococcus aureus is acknowledged as one of the important pathogens isolated from humans and animals. However, the S. aureus causing severe respiratory diseases in rabbits have not been well characterised. A S. aureus named FZHW001, isolated from the lungs of dead rabbits with severe respiratory disease, was characterised by artificial infection of rabbits, detection of virulence factors, multi-locus sequencing typing and antimicrobial susceptibility test. The FZHW001 infected rabbits showed identical respiratory symptoms to those of naturally infected ones, and the isolate could spread through directed contact among rabbits. The isolate was typed into clonal complex 121 and carried 7 of 13 tested virulence factors. Furthermore, the isolate was identified to be methicillin-susceptible S. aureus and was susceptible to 7 of 12 tested antibiotics. This study first describes the characteristics of S. aureus isolated from rabbits causing severe respiratory disease, which will help in further understanding the pathogenic mechanisms of S. aureus in rabbits., This work was supported by the Outstanding Youth Fund of Fujian Academy of Agricultural Sciences (JC2018-1) and National Rabbit Industry Technology System of People’s Republic of China (CARS-43-G-5).

Published
2019

41. Characterisation of Bordetella bronchiseptica isolated from rabbits in Fujian, China

Author

Xie Xiping, Sun Shikun, Wang Jinxiang, Chen Dongjin, Chen Yanfeng, and Sang Lei

Subjects
0301 basic medicine, China, 040301 veterinary sciences, Epidemiology, Respiratory Tract Diseases, rabbit, Virulence, Biology, Bordetella bronchiseptica, Antimicrobial susceptibility, multi-locus sequence typing, law.invention, Microbiology, 0403 veterinary science, 03 medical and health sciences, law, Ampicillin, Ceftizoxime, Drug Resistance, Bacterial, medicine, Animals, Typing, Polymerase chain reaction, Phylogeny, Bordetella Infections, Original Paper, virulence genes, Genetic Variation, 04 agricultural and veterinary sciences, biology.organism_classification, Anti-Bacterial Agents, 030104 developmental biology, Infectious Diseases, Ceftriaxone, Rabbits, Cefixime, medicine.drug
Abstract

Bordetella bronchiseptica is a potential zoonotic pathogen, which mainly causes respiratory diseases in humans and a variety of animal species. B. bronchiseptica is one of the important pathogens isolated from rabbits in Fujian Province. However, the knowledge of the epidemiology and characteristics of the B. bronchiseptica in rabbits in Fujian Province is largely unknown. In this study, 219 B. bronchiseptica isolates recovered from lung samples of dead rabbits with respiratory diseases in Fujian Province were characterised by multi-locus sequencing typing, screening virulence genes and testing antimicrobial susceptibility. The results showed that the 219 isolates were typed into 11 sequence types (STs) including five known STs (ST6, ST10, ST12, ST14 and ST33) and six new STs (ST88, ST89, ST90, ST91, ST92 and ST93) and the ST33 (30.14%, 66/219), ST14 (26.94%, 59/219) and ST12 (16.44%, 36/219) were the three most prevalent STs. Surprisingly, all the 219 isolates carried the five virulence genes (fhaB, prn, cyaA, dnt and bteA) in the polymerase chain reaction screening. Moreover, the isolates were resistant to cefixime, ceftizoxime, cefatriaxone and ampicillin at rates of 33.33%, 31.05%, 11.87% and 3.20%, respectively. This study showed the genetic diversity of B. bronchiseptica in rabbits in Fujian Province, and the colonisation of the human-associated ST12 strain in rabbits in Fujian Province. The results might be useful for monitoring the epidemic strains, developing preventive methods and preventing the transmission of epidemic strains from rabbits to humans.

Published
2020

44. Characterization of Pasteurella multocida isolated from dead rabbits with respiratory disease in Fujian, China

Author

Wang Jinxiang, Chen Yanfeng, Chen Dongjin, Sun Shikun, Sang Lei, and Xie Xiping

Subjects
Serotype, China, Pasteurella multocida, medicine.drug_class, Antibiotics, Pasteurella Infections, Virulence, Rabbit, Antimicrobial susceptibility, Microbiology, Multi-locus sequence typing, Genotype, medicine, Prevalence, Animals, Pathogen, lcsh:Veterinary medicine, General Veterinary, biology, Virulence factors, Kanamycin, General Medicine, Gene Expression Regulation, Bacterial, biology.organism_classification, Bacterial Typing Techniques, Streptomycin, lcsh:SF600-1100, Rabbits, medicine.drug, Research Article
Abstract

Background Pasteurella multocida is one of the important pathogens that infect rabbits, causing major economic losses in commercial rabbit farming. In this study, 205 P. multocida isolates recovered from lungs of dead rabbits with respiratory disease were defined by capsular serogroups, lipopolysaccharide (LPS) genotypes, multi-locus sequence types and screened virulence factors by using PCR assays, and tested antimicrobial susceptibility. Results The 205 isolates were assigned into 2 capsular types, A and D, and 2 LPS genotypes, L3 and L6. When combining capsular types with LPS genotypes, 4 serotypes were detected. A:L3 (51.22%, 105/205) was the most predominant serotype, followed by A:L6 (24.88%, 51/205), D:L6 (19.02%, 39/205) and D:L3 (4.88%, 10/205). The 205 isolates were grouped into 3 sequence types, ST10, ST11 and ST12. ST12 (56.10%, 115/205) was the most prevalent sequence type, followed by ST10 (24.88%, 51/205) and ST11 (19.02%, 39/205). In the 205 isolates, virulence associated genes ptfA, fur, hgbB, ompA, ompH and oma87 were positive in the PCR screening, whereas the toxA and tbpA genes were negative. Notably, the 156 capsular serogroup A isolates carried the pmHAS gene. All the 205 isolates were susceptible to most of the used antibiotics, except for streptomycin, gentamycin, kanamycin and ceftriaxone, and the resistance rates of which were 27.80, 15.61, 9.27 and 2.44%, respectively. Conclusions This study, for the first time, described the prevalence and characteristics of P. multocida causing respiratory disease in rabbits in Fujian Province, which might be useful for tracking the epidemic strains and development of efficient vaccines and methods to prevent and control the pathogen.

Published
2019

45. Characterisation ofStaphylococcus aureusisolated from rabbits in Fujian, China

Author

Sang Lei, Xie Xiping, Chen Dongjin, Sun Shikun, Wang Jinxiang, and Yunbo Chen

Subjects
0301 basic medicine, Epidemiology, medicine.drug_class, 030106 microbiology, Antibiotics, Virulence, Kanamycin, Biology, medicine.disease_cause, Azithromycin, Microbiology, Penicillin, 03 medical and health sciences, 030104 developmental biology, Infectious Diseases, Streptomycin, Staphylococcus aureus, medicine, Typing, medicine.drug
Abstract

Staphylococcus aureushas been recognised as one of the important zoonotic pathogens. However, knowledge about the epidemiology and genetic characteristics ofS. aureusin rabbits was limited. The aim of this study was to determine the characteristics of 281S. aureusisolated from dead rabbits of nine rabbit farms in Fujian Province, China. All the isolates were characterised by multi-locus sequencing typing, detection of virulence factors and antimicrobial susceptibility test. The results showed that the 281 isolates were grouped into two sequence types, ST121 (13.52%, 38/281) and ST398 (86.48%, 243/281). Surprisingly, the ST121 strains were only recovered from the lung samples from one of the nine rabbit farms studied. In the 281 isolates, the virulence genes ofnuc,hla,hlb,clfA,clfBandfnbpAwere positive, whereas thesea,seb,tsst,etaandetbgenes were negative. Notably, the 38 ST121 isolates carried thepvlgene. All the 281 isolates were methicillin-susceptibleS. aureus, and the isolates were susceptible to most of the used antibiotics, except for streptomycin, kanamycin, azithromycin and penicillin, and the resistance rates of which were 23.84%, 19.57%, 16.01% and 11.03%, respectively. This study first described the epidemiology and characteristics ofS. aureusin rabbits in Fujian Province, which will help in tracking the evolution of epidemic strains and preventing the rabbit–human transmission events.

Published
2019

47. Characterisation of Staphylococcus aureus strain causing severe respiratory disease in rabbits

Author

Fujian Academy of Agricultural Sciences, China, Earmarked Fund for Modern Agro-industry Technology Research System, China, Wang, Jinxiang, Sang, Lei, Chen, Yanfeng, Sun, Shikun, Chen, Dongjin, Xie, Xiping, Fujian Academy of Agricultural Sciences, China, Earmarked Fund for Modern Agro-industry Technology Research System, China, Wang, Jinxiang, Sang, Lei, Chen, Yanfeng, Sun, Shikun, Chen, Dongjin, and Xie, Xiping

Abstract

[EN] Staphylococcus aureus is acknowledged as one of the important pathogens isolated from humans and animals. However, the S. aureus causing severe respiratory diseases in rabbits have not been well characterised. A S. aureus named FZHW001, isolated from the lungs of dead rabbits with severe respiratory disease, was characterised by artificial infection of rabbits, detection of virulence factors, multi-locus sequencing typing and antimicrobial susceptibility test. The FZHW001 infected rabbits showed identical respiratory symptoms to those of naturally infected ones, and the isolate could spread through directed contact among rabbits. The isolate was typed into clonal complex 121 and carried 7 of 13 tested virulence factors. Furthermore, the isolate was identified to be methicillin-susceptible S. aureus and was susceptible to 7 of 12 tested antibiotics. This study first describes the characteristics of S. aureus isolated from rabbits causing severe respiratory disease, which will help in further understanding the pathogenic mechanisms of S. aureus in rabbits.

Published
2019

48. Advances and perspectives of research on soil moisture response to afforestation: a review.

Author

Kong, Wei, Gao, Fei, Sun, Shikun, Lei, Yuhan, Gu, Jiali, and Luan, Xiaobo

Subjects
*SOIL productivity, *SOIL moisture, *AFFORESTATION, *SOILS, *HETEROGENEITY
Abstract

The soil moisture (SM) balance is affected by afforestation. In this study, the CiteSpace software was used to review the literature on the impact of afforestation on SM, and we further present research findings on the Loess Plateau (LP) as follows: (1) the spatio-temporal heterogeneity of SM varies depending on the research scale; (2) the presence of a dried soil layer significantly hinders deep SM recharge; (3) the accumulation of net primary productivity (NPP) may be occurring at the expense of SM. The following research topics are discussed: (1) the recharge and consumption of deep SM in LP; (2) the green water footprint and its shortage of non-crop greenery for evaluating SM allocation and use; (3) the development of inclusive models encompassing different spatial scales to assess the limits of soil water carrying capacity for vegetation (SWCCV). [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results