8 results on '"Qi, Jiyang"'
Search Results
2. Occluded Video Instance Segmentation: A Benchmark
- Author
-
Qi, Jiyang, Gao, Yan, Hu, Yao, Wang, Xinggang, Liu, Xiaoyu, Bai, Xiang, Belongie, Serge, Yuille, Alan, Torr, Philip H. S., and Bai, Song
- Published
- 2022
- Full Text
- View/download PDF
3. You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detection
- Author
-
Fang, Yuxin, Liao, Bencheng, Wang, Xinggang, Fang, Jiemin, Qi, Jiyang, Wu, Rui, Niu, Jianwei, and Liu, Wenyu
- Subjects
FOS: Computer and information sciences ,Computer Science - Machine Learning ,Artificial Intelligence (cs.AI) ,Computer Science - Artificial Intelligence ,Computer Vision and Pattern Recognition (cs.CV) ,Computer Science - Computer Vision and Pattern Recognition ,Machine Learning (cs.LG) - Abstract
Can Transformer perform 2D object- and region-level recognition from a pure sequence-to-sequence perspective with minimal knowledge about the 2D spatial structure? To answer this question, we present You Only Look at One Sequence (YOLOS), a series of object detection models based on the vanilla Vision Transformer with the fewest possible modifications, region priors, as well as inductive biases of the target task. We find that YOLOS pre-trained on the mid-sized ImageNet-1k dataset only can already achieve quite competitive performance on the challenging COCO object detection benchmark, e.g., YOLOS-Base directly adopted from BERT-Base architecture can obtain 42.0 box AP on COCO val. We also discuss the impacts as well as limitations of current pre-train schemes and model scaling strategies for Transformer in vision through YOLOS. Code and pre-trained models are available at https://github.com/hustvl/YOLOS., NeurIPS 2021 Camera Ready
- Published
- 2021
4. The stereo matching algorithm based on an improved adaptive support window.
- Author
-
Qi, Jiyang and Liu, Liang
- Subjects
- *
ALGORITHMS , *CENSUS , *PIXELS , *COLOR , *IMMUNITY , *NOISE - Abstract
In binocular stereo matching, there has been a problem of low matching accuracy and noise immunity in discontinuous regions and weak‐textured regions. This paper proposes a stereo matching algorithm based on an improved adaptive support window. In the cost computation stage, first, according to the preset arm length and colour threshold, a cross‐based arm is obtained, which centres on the pixel to be matched; then the adaptive regions of the vertical arm and the horizontal arm are constructed respectively, which have different shape and size. Finally, the union of the two adaptive regions is used as the final support window of Census transform. Performance evaluations on Middlebury stereo data sets demonstrate that the proposed algorithm outperforms other seven most challenging stereo matching algorithms. The mismatching rate of this algorithm is greatly reduced, and the anti‐noise performance is also improved considerably. Because the construction of the adaptive region is based on strict criteria and comprehensive consideration, the algorithm proposed in the paper can improve the matching accuracy in the weak‐textured regions and discontinuous disparity regions. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
5. Multi-Granularity Modularized Network for Abstract Visual Reasoning
- Author
-
Tang, Xiangru, Wang, Haoyuan, Pan, Xiang, and Qi, Jiyang
- Subjects
FOS: Computer and information sciences ,Artificial Intelligence (cs.AI) ,Computer Science - Artificial Intelligence ,Computer Vision and Pattern Recognition (cs.CV) ,Computer Science - Computer Vision and Pattern Recognition - Abstract
visual reasoning connects mental abilities to the physical world, which is a crucial factor in cognitive development. Most toddlers display sensitivity to this skill, but it is not easy for machines. Aimed at it, we focus on the Raven Progressive Matrices Test, designed to measure cognitive reasoning. Recent work designed some black-boxes to solve it in an end-to-end fashion, but they are incredibly complicated and difficult to explain. Inspired by cognitive studies, we propose a Multi-Granularity Modularized Network (MMoN) to bridge the gap between the processing of raw sensory information and symbolic reasoning. Specifically, it learns modularized reasoning functions to model the semantic rule from the visual grounding in a neuro-symbolic and semi-supervision way. To comprehensively evaluate MMoN, our experiments are conducted on the dataset of both seen and unseen reasoning rules. The result shows that MMoN is well suited for abstract visual reasoning and also explainable on the generalization test.
- Published
- 2020
6. A Study of Drag Reduction on Cylinders with Different V-Groove Depths on the Surface.
- Author
-
Qi, Jiyang, Qi, Yue, Chen, Qunyan, and Yan, Fei
- Subjects
COMPUTATIONAL fluid dynamics ,DRAG (Aerodynamics) ,DRAG reduction ,PARTICLE image velocimetry ,REYNOLDS number ,FLOW velocity ,DRAG coefficient - Abstract
In this study, the drag reduction effect is studied for a cylinder with different V-groove depths on its surface using a k-ω/SST (Shear Stress Transport) turbulence model of computational fluid dynamics (CFD), while a particle image velocimetry (PIV) system is employed to analyze the wake characteristics for a smooth cylinder and a cylinder with different V-groove depths on its surface at different Reynolds numbers. The study focuses on the characteristics of the different V-groove depths on lift coefficient, drag coefficient, the velocity distribution of flow field, pressure coefficient, vortex shedding, and vortex structure. In comparison with a smooth cylinder, the lift coefficient and drag coefficient can be reduced for a cylinder with different V-groove depths on its surface, and the maximum reduction rates of lift coefficient and drag coefficient are about 34.4% and 16%, respectively. Otherwise, the vortex structure presents a complete symmetry for the smooth cylinder, however, the symmetry of the vortex structure becomes insignificant for the V-shaped groove structure with different depths. This is also an important reason for the drag reduction effect of a cylinder with a V-groove surface. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
7. Liquid–Solid Flow Characteristics in Vertical Swirling Hydraulic Transportation with Tangential Jet Inlet.
- Author
-
Qi, Jiyang, Yin, Jie, Yan, Fei, Liu, Ping, Wang, Tieli, and Chen, Chen
- Subjects
DRAG force ,DRAG coefficient ,PIPELINE transportation ,SWIRLING flow ,VORTEX motion ,INLETS - Abstract
In order to improve the efficiency and safety of vertical hydraulic transport systems for non-spherical particles, a new pipeline transport system with a tangential jet inlet is adopted in this study, and a modified non-spherical drag coefficient model is used to analyze the liquid–solid flow characteristics based on the CFD-DEM (Computational Fluid Dynamics-Discrete Element Method) coupling method. The focus of the study is on the influence of different tangential flow proportions in terms of the velocity distribution, the vorticity, the total pressure, the concentration and drag force of particles of various shapes. The conveying efficiency is measured according to the fluid velocity distribution and the particle concentration, and the safety of conveying is evaluated according to the flow structure and drag force of the particles. The result shows that the velocity of the swirling pipes is significantly higher than the straight pipe. With the increase of the tangential flow proportion, the swirling number and the vorticity magnitude increase, and the vortex core is broken and merged more quickly. Furthermore, the concentration gap and axial drag force gap between particles of various shapes are reduced with the effect of swirling flow, the particle concentration increases, and the particles of each component are uniformly mixed and transported. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
8. Dynamic Analysis of Wake Characteristics of the Circular Cylinder with a Dimpled Surface.
- Author
-
Qi, Jiyang, Shao, Yuyao, Chen, Qunyan, Liu, Ping, Chen, Chen, Wang, Deqiang, Yang, Haifeng, and Yan, Fei
- Subjects
VORTEX shedding ,PROPER orthogonal decomposition ,PARTICLE image velocimetry ,FLOW velocity ,KINETIC energy ,REYNOLDS stress ,STRESS concentration - Abstract
In order to understand the wake characteristics for the circular cylinder with the dimpled structure, particle image velocimetry (PIV) and proper orthogonal decomposition (POD) method are used to measure and analyze the flow field velocity and the flow characteristic of the wake flow of the circular cylinder with the dimpled surface. This study focuses on the distribution of Reynolds stress and turbulent kinetic energy, the velocity profiles and recirculation zones, and the velocity fluctuating characteristics of flow field with POD technology. It is found that the equivalent high-intensity Reynolds stress and turbulent kinetic energy regions of the circular cylinder with the dimpled surface are smaller, and the peak values are lower, and the velocity gradient in the wake region of the circular cylinder with the dimpled surface is larger. Otherwise, the energy contained by the dominant modes of the smooth cylinder is larger than that contained by the dominant modes of the circular cylinders with the dimpled surface, which means the energy of the dimpled cylinder is more distributed. At the same time, it is observed that the dimpled structure will decrease the vortex shedding intensity, but may increase the vortex shedding frequency, and destroy the inherent flow mode of the flow field around the cylinder. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.