Search

A unified theoretical framework is suggested to examine the energy dissipation properties at all stages of additive implicit-explicit Runge-Kutta (IERK) methods up to fourth-order accuracy for gradient flow problems. We construct some parameterized IERK methods by applying the so-called first same as last method, that is, the diagonally implicit Runge-Kutta method with the explicit first stage and stiffly-accurate assumption for the linear stiff term, and applying the explicit Runge-Kutta method for the nonlinear term. The main part of the novel framework is to construct the differential forms and the associated differentiation matrices of IERK methods by using the difference coefficients of method and the so-called discrete orthogonal convolution kernels. As the main result, we prove that an IERK method can preserve the original energy dissipation law unconditionally if the associated differentiation matrix is positive semi-definite. The recent indicator, namely average energy dissipation rate, is also adopted for these multi-stage methods to evaluate the overall energy dissipation rate of an IERK method such that one can choose proper parameters in some parameterized IERK methods. It is found that the selection of method parameters in the IERK methods is at least as important as the selection of different IERK methods. Extensive numerical experiments are also included to support our theory., Comment: 32 pages, 3 tables, 40 figures

Explicit integrating factor Runge-Kutta methods are attractive and popular in developing high-order maximum bound principle preserving time-stepping schemes for Allen-Cahn type gradient flows. However, they always suffer from the non-preservation of steady-state solution and original energy dissipation law. To overcome these disadvantages, some new integrating factor methods are developed by using two classes of difference correction, including the telescopic correction and nonlinear-term translation correction, enforcing the preservation of steady-state solution. Then the original energy dissipation properties of the new methods are examined by using the associated differential forms and the differentiation matrices. As applications, some new integrating factor Runge-Kutta methods up to third-order maintaining the original energy dissipation law are constructed by applying the difference correction strategies to some popular explicit integrating factor methods in the literature. Extensive numerical experiments are presented to support our theory and to demonstrate the improved performance of new methods., Comment: 25 pages,55 figures

This is a comprehensive review of the YOLO series of systems. Different from previous literature surveys, this review article re-examines the characteristics of the YOLO series from the latest technical point of view. At the same time, we also analyzed how the YOLO series continued to influence and promote real-time computer vision-related research and led to the subsequent development of computer vision and language models.We take a closer look at how the methods proposed by the YOLO series in the past ten years have affected the development of subsequent technologies and show the applications of YOLO in various fields. We hope this article can play a good guiding role in subsequent real-time computer vision development., Comment: 13 pages, 14 figures

We propose a unified theoretical framework to examine the energy dissipation properties at all stages of explicit exponential Runge-Kutta (EERK) methods for gradient flow problems. The main part of the novel framework is to construct the differential form of EERK method by using the difference coefficients of method and the so-called discrete orthogonal convolution kernels. As the main result, we prove that an EERK method can preserve the original energy dissipation law unconditionally if the associated differentiation matrix is positive semi-definite. A simple indicator, namely average dissipation rate, is also introduced for these multi-stage methods to evaluate the overall energy dissipation rate of an EERK method such that one can choose proper parameters in some parameterized EERK methods or compare different kinds of EERK methods. Some existing EERK methods in the literature are evaluated from the perspective of preserving the original energy dissipation law and the energy dissipation rate. Some numerical examples are also included to support our theory., Comment: 35 pages, 44 figures

Today's deep learning methods focus on how to design the most appropriate objective functions so that the prediction results of the model can be closest to the ground truth. Meanwhile, an appropriate architecture that can facilitate acquisition of enough information for prediction has to be designed. Existing methods ignore a fact that when input data undergoes layer-by-layer feature extraction and spatial transformation, large amount of information will be lost. This paper will delve into the important issues of data loss when data is transmitted through deep networks, namely information bottleneck and reversible functions. We proposed the concept of programmable gradient information (PGI) to cope with the various changes required by deep networks to achieve multiple objectives. PGI can provide complete input information for the target task to calculate objective function, so that reliable gradient information can be obtained to update network weights. In addition, a new lightweight network architecture -- Generalized Efficient Layer Aggregation Network (GELAN), based on gradient path planning is designed. GELAN's architecture confirms that PGI has gained superior results on lightweight models. We verified the proposed GELAN and PGI on MS COCO dataset based object detection. The results show that GELAN only uses conventional convolution operators to achieve better parameter utilization than the state-of-the-art methods developed based on depth-wise convolution. PGI can be used for variety of models from lightweight to large. It can be used to obtain complete information, so that train-from-scratch models can achieve better results than state-of-the-art models pre-trained using large datasets, the comparison results are shown in Figure 1. The source codes are at: https://github.com/WongKinYiu/yolov9.

We provide a new theoretical framework for the variable-step deferred correction (DC) methods based on the well-known BDF2 formula. By using the discrete orthogonal convolution kernels, some high-order BDF2-DC methods are proven to be stable on arbitrary time grids according to the recent definition of stability (SINUM, 60: 2253-2272). It significantly relaxes the existing step-ratio restrictions for the BDF2-DC methods (BIT, 62: 1789-1822). The associated sharp error estimates are established by taking the numerical effects of the starting approximations into account, and they suggest that the BDF2-DC methods have no aftereffect, that is, the lower-order starting scheme for the BDF2 scheme will not cause a loss in the accuracy of the high-order BDF2-DC methods. Extensive tests on the graded and random time meshes are presented to support the new theory., Comment: 27 pages, 12 tables, 8 figures

We build an asymptotically compatible energy of the variable-step L2-$1_{\sigma}$ scheme for the time-fractional Allen-Cahn model with the Caputo's fractional derivative of order $\alpha\in(0,1)$, under a weak step-ratio constraint $\tau_k/\tau_{k-1}\geq r_{\star}(\alpha)$ for $k\ge2$, where $\tau_k$ is the $k$-th time-step size and $r_{\star}(\alpha)\in(0.3865,0.4037)$ for $\alpha\in(0,1)$. It provides a positive answer to the open problem in [J. Comput. Phys., 414:109473], and, to the best of our knowledge, it is the first second-order nonuniform time-stepping scheme to preserve both the maximum bound principle and the energy dissipation law of time-fractional Allen-Cahn model. The compatible discrete energy is constructed via a novel discrete gradient structure of the second-order L2-$1_{\sigma}$ formula by a local-nonlocal splitting technique. It splits the discrete fractional derivative into two parts: one is a local term analogue to the trapezoid rule of the first derivative and the other is a nonlocal summation analogue to the L1 formula of Caputo derivative. Numerical examples with an adaptive time-stepping strategy are provided to show the effectiveness of our scheme and the asymptotic properties of the associated modified energy., Comment: 21 pages,23 figues

Multi-task learning (MTL) aims to learn multiple tasks using a single model and jointly improve all of them assuming generalization and shared semantics. Reducing conflicts between tasks during joint learning is difficult and generally requires careful network design and extremely large models. We propose building on You Only Learn One Representation (YOLOR), a network architecture specifically designed for multitasking. YOLOR leverages both explicit and implicit knowledge, from data observations and learned latents, respectively, to improve a shared representation while minimizing the number of training parameters. However, YOLOR and its follow-up, YOLOv7, only trained two tasks at once. In this paper, we jointly train object detection, instance segmentation, semantic segmentation, and image captioning. We analyze tradeoffs and attempt to maximize sharing of semantic information. Through our architecture and training strategies, we find that our method achieves competitive performance on all tasks while maintaining a low parameter count and without any pre-training. We will release code soon.

The discrete gradient structure and the positive definiteness of discrete fractional integrals or derivatives are fundamental to the numerical stability in long-time simulation of nonlinear integro-differential models. We build up a discrete gradient structure for a class of second-order variable-step approximations of fractional Riemann-Liouville integral and fractional Caputo derivative. Then certain variational energy dissipation laws at discrete levels of the corresponding variable-step Crank-Nicolson type methods are established for time-fractional Allen-Cahn and time-fractional Klein-Gordon type models. They are shown to be asymptotically compatible with the associated energy laws of the classical Allen-Cahn and Klein-Gordon equations in the associated fractional order limits.Numerical examples together with an adaptive time-stepping procedure are provided to demonstrate the effectiveness of our second-order methods., Comment: 25 pages, 16 figures, 2 tables

We propose a post-processor, called NeighborTrack, that leverages neighbor information of the tracking target to validate and improve single-object tracking (SOT) results. It requires no additional data or retraining. Instead, it uses the confidence score predicted by the backbone SOT network to automatically derive neighbor information and then uses this information to improve the tracking results. When tracking an occluded target, its appearance features are untrustworthy. However, a general siamese network often cannot tell whether the tracked object is occluded by reading the confidence score alone, because it could be misled by neighbors with high confidence scores. Our proposed NeighborTrack takes advantage of unoccluded neighbors' information to reconfirm the tracking target and reduces false tracking when the target is occluded. It not only reduces the impact caused by occlusion, but also fixes tracking problems caused by object appearance changes. NeighborTrack is agnostic to SOT networks and post-processing methods. For the VOT challenge dataset commonly used in short-term object tracking, we improve three famous SOT networks, Ocean, TransT, and OSTrack, by an average of ${1.92\%}$ EAO and ${2.11\%}$ robustness. For the mid- and long-term tracking experiments based on OSTrack, we achieve state-of-the-art ${72.25\%}$ AUC on LaSOT and ${75.7\%}$ AO on GOT-10K. Code duplication can be found in https://github.com/franktpmvu/NeighborTrack., Comment: This paper was accepted by 9th International Workshop on Computer Vision in Sports (CVsports) 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

Designing a high-efficiency and high-quality expressive network architecture has always been the most important research topic in the field of deep learning. Most of today's network design strategies focus on how to integrate features extracted from different layers, and how to design computing units to effectively extract these features, thereby enhancing the expressiveness of the network. This paper proposes a new network design strategy, i.e., to design the network architecture based on gradient path analysis. On the whole, most of today's mainstream network design strategies are based on feed forward path, that is, the network architecture is designed based on the data path. In this paper, we hope to enhance the expressive ability of the trained model by improving the network learning ability. Due to the mechanism driving the network parameter learning is the backward propagation algorithm, we design network design strategies based on back propagation path. We propose the gradient path design strategies for the layer-level, the stage-level, and the network-level, and the design strategies are proved to be superior and feasible from theoretical analysis and experiments., Comment: 12 pages, 9 figures

The paper presents a new method, SearchTrack, for multiple object tracking and segmentation (MOTS). To address the association problem between detected objects, SearchTrack proposes object-customized search and motion-aware features. By maintaining a Kalman filter for each object, we encode the predicted motion into the motion-aware feature, which includes both motion and appearance cues. For each object, a customized fully convolutional search engine is created by SearchTrack by learning a set of weights for dynamic convolutions specific to the object. Experiments demonstrate that our SearchTrack method outperforms competitive methods on both MOTS and MOT tasks, particularly in terms of association accuracy. Our method achieves 71.5 HOTA (car) and 57.6 HOTA (pedestrian) on the KITTI MOTS and 53.4 HOTA on MOT17. In terms of association accuracy, our method achieves state-of-the-art performance among 2D online methods on the KITTI MOTS. Our code is available at https://github.com/qa276390/SearchTrack., Comment: BMVC 2022. Code: https://github.com/qa276390/SearchTrack

A new discrete energy dissipation law of the variable-step fractional BDF2 (second-order backward differentiation formula) scheme is established for time-fractional Cahn-Hilliard model with the Caputo's fractional derivative of order $\alpha\in(0,1)$, under a weak step-ratio constraint $0.4753\le \tau_k/\tau_{k-1}

Chemical transport models (CTMs) have been widely used to provide instructions for the control of ozone (O3) pollution. However, we find large discrepancies between observation- and model-based urban O3 chemical regimes: volatile organic compound (VOC)-limited regimes over N. China and weak nitrogen oxides (NOx)-limited regimes over S. China in observations, in contrast to simulations with widespread distributions of strong NOx-limited regimes. The conflicting O3 evolutions are caused by underestimated urban NOx concentrations and the possible overestimation of biogenic VOC emissions. Reductions in NOx emissions, in response to regulations, have thus led to an unintended deterioration of O3 pollution over N. China provinces, for example, an increase in surface O3 by approximately 7 ppb over the Sichuan Basin (SCB) in 2014-2020. The NOx-induced urban O3 changes resulted in an increase in premature mortality by approximately 3000 cases in 2015-2020.

Coarse particulate matter (PM) is a serious air pollution problem in East Asia. Analysis of air quality network observations in the North China Plain and the Seoul Metropolitan Area shows that it is mainly anthropogenic and has decreased by 21% over 2015-2019. This anthropogenic coarse PM is generally not included in air quality models but scavenges nitric acid to suppress the formation of fine particulate (PM2.5) nitrate, a major contributor to PM2.5 pollution. Including it in the GEOS-Chem model decreases simulated PM2.5 nitrate to improve agreement with observations. Decreasing anthropogenic coarse PM over 2015-2019 directly increases PM2.5 nitrate in summer, offsetting the effect of other emission controls, while in winter it increases the sensitivity of PM2.5 nitrate to ammonia and sulfur dioxide emissions. Our work implies the need for stronger ammonia and nitrogen oxides emission controls to improve PM2.5 air quality as coarse PM continues to decrease., Comment: We have revised this work considerably and want to avoid the confusion

YOLOv7 surpasses all known object detectors in both speed and accuracy in the range from 5 FPS to 160 FPS and has the highest accuracy 56.8% AP among all known real-time object detectors with 30 FPS or higher on GPU V100. YOLOv7-E6 object detector (56 FPS V100, 55.9% AP) outperforms both transformer-based detector SWIN-L Cascade-Mask R-CNN (9.2 FPS A100, 53.9% AP) by 509% in speed and 2% in accuracy, and convolutional-based detector ConvNeXt-XL Cascade-Mask R-CNN (8.6 FPS A100, 55.2% AP) by 551% in speed and 0.7% AP in accuracy, as well as YOLOv7 outperforms: YOLOR, YOLOX, Scaled-YOLOv4, YOLOv5, DETR, Deformable DETR, DINO-5scale-R50, ViT-Adapter-B and many other object detectors in speed and accuracy. Moreover, we train YOLOv7 only on MS COCO dataset from scratch without using any other datasets or pre-trained weights. Source code is released in https://github.com/WongKinYiu/yolov7.

The convergence of variable-step L1 scheme is studied for the time-fractional molecular beam epitaxy (MBE) model with slope selection.A novel asymptotically compatible $L^2$ norm error estimate of the variable-step L1 scheme is established under a convergence-solvability-stability (CSS)-consistent time-step constraint. The CSS-consistent condition means that the maximum step-size limit required for convergence is of the same order to that for solvability and stability (in certain norms) as the small interface parameter $\epsilon\rightarrow 0^+$. To the best of our knowledge, it is the first time to establish such error estimate for nonlinear subdiffusion problems. The asymptotically compatible convergence means that the error estimate is compatible with that of backward Euler scheme for the classical MBE model as the fractional order $\alpha\rightarrow 1^-$. Just as the backward Euler scheme can maintain the physical properties of the MBE equation, the variable-step L1 scheme can also preserve the corresponding properties of the time-fractional MBE model, including the volume conservation, variational energy dissipation law and $L^2$ norm boundedness. Numerical experiments are presented to support our theoretical results., Comment: 23 pages,21 figures,4 tables

This is one of our series works on discrete energy analysis of the variable-step BDF schemes. In this part, we present stability and convergence analysis of the third-order BDF (BDF3) schemes with variable steps for linear diffusion equations, see e.g. [SIAM J. Numer. Anal., 58:2294-2314] and [Math. Comp., 90: 1207-1226] for our previous works on the BDF2 scheme. To this aim, we first build up a discrete gradient structure of the variable-step BDF3 formula under the condition that the adjacent step ratios are less than 1.4877, by which we can establish a discrete energy dissipation law. Mesh-robust stability and convergence analysis in the $L^2$ norm are then obtained. Here the mesh robustness means that the solution errors are well controlled by the maximum time-step size but independent of the adjacent time-step ratios. We also present numerical tests to support our theoretical results., Comment: 21 pages, 5 tables

Learning to capture human motion is essential to 3D human pose and shape estimation from monocular video. However, the existing methods mainly rely on recurrent or convolutional operation to model such temporal information, which limits the ability to capture non-local context relations of human motion. To address this problem, we propose a motion pose and shape network (MPS-Net) to effectively capture humans in motion to estimate accurate and temporally coherent 3D human pose and shape from a video. Specifically, we first propose a motion continuity attention (MoCA) module that leverages visual cues observed from human motion to adaptively recalibrate the range that needs attention in the sequence to better capture the motion continuity dependencies. Then, we develop a hierarchical attentive feature integration (HAFI) module to effectively combine adjacent past and future feature representations to strengthen temporal correlation and refine the feature representation of the current frame. By coupling the MoCA and HAFI modules, the proposed MPS-Net excels in estimating 3D human pose and shape in the video. Though conceptually simple, our MPS-Net not only outperforms the state-of-the-art methods on the 3DPW, MPI-INF-3DHP, and Human3.6M benchmark datasets, but also uses fewer network parameters. The video demos can be found at https://mps-net.github.io/MPS-Net/., Comment: Accepted by CVPR 2022