Search

Diffusion models for garment-centric human generation from text or image prompts have garnered emerging attention for their great application potential. However, existing methods often face a dilemma: lightweight approaches, such as adapters, are prone to generate inconsistent textures; while finetune-based methods involve high training costs and struggle to maintain the generalization capabilities of pretrained diffusion models, limiting their performance across diverse scenarios. To address these challenges, we propose DreamFit, which incorporates a lightweight Anything-Dressing Encoder specifically tailored for the garment-centric human generation. DreamFit has three key advantages: (1) \textbf{Lightweight training}: with the proposed adaptive attention and LoRA modules, DreamFit significantly minimizes the model complexity to 83.4M trainable parameters. (2)\textbf{Anything-Dressing}: Our model generalizes surprisingly well to a wide range of (non-)garments, creative styles, and prompt instructions, consistently delivering high-quality results across diverse scenarios. (3) \textbf{Plug-and-play}: DreamFit is engineered for smooth integration with any community control plugins for diffusion models, ensuring easy compatibility and minimizing adoption barriers. To further enhance generation quality, DreamFit leverages pretrained large multi-modal models (LMMs) to enrich the prompt with fine-grained garment descriptions, thereby reducing the prompt gap between training and inference. We conduct comprehensive experiments on both $768 \times 512$ high-resolution benchmarks and in-the-wild images. DreamFit surpasses all existing methods, highlighting its state-of-the-art capabilities of garment-centric human generation., Comment: Accepted at AAAI 2025

Most existing mobile robotic datasets primarily capture static scenes, limiting their utility for evaluating robotic performance in dynamic environments. To address this, we present a mobile robot oriented large-scale indoor dataset, denoted as THUD++ (TsingHua University Dynamic) robotic dataset, for dynamic scene understanding. Our current dataset includes 13 large-scale dynamic scenarios, combining both real-world and synthetic data collected with a real robot platform and a physical simulation platform, respectively. The RGB-D dataset comprises over 90K image frames, 20M 2D/3D bounding boxes of static and dynamic objects, camera poses, and IMU. The trajectory dataset covers over 6,000 pedestrian trajectories in indoor scenes. Additionally, the dataset is augmented with a Unity3D-based simulation platform, allowing researchers to create custom scenes and test algorithms in a controlled environment. We evaluate state-of-the-art methods on THUD++ across mainstream indoor scene understanding tasks, e.g., 3D object detection, semantic segmentation, relocalization, pedestrian trajectory prediction, and navigation. Our experiments highlight the challenges mobile robots encounter in indoor environments, especially when navigating in complex, crowded, and dynamic scenes. By sharing this dataset, we aim to accelerate the development and testing of mobile robot algorithms, contributing to real-world robotic applications., Comment: arXiv admin note: substantial text overlap with arXiv:2406.19791

The reconstruction of indoor scenes remains challenging due to the inherent complexity of spatial structures and the prevalence of textureless regions. Recent advancements in 3D Gaussian Splatting have improved novel view synthesis with accelerated processing but have yet to deliver comparable performance in surface reconstruction. In this paper, we introduce 2DGS-Room, a novel method leveraging 2D Gaussian Splatting for high-fidelity indoor scene reconstruction. Specifically, we employ a seed-guided mechanism to control the distribution of 2D Gaussians, with the density of seed points dynamically optimized through adaptive growth and pruning mechanisms. To further improve geometric accuracy, we incorporate monocular depth and normal priors to provide constraints for details and textureless regions respectively. Additionally, multi-view consistency constraints are employed to mitigate artifacts and further enhance reconstruction quality. Extensive experiments on ScanNet and ScanNet++ datasets demonstrate that our method achieves state-of-the-art performance in indoor scene reconstruction.

The density limit of lower hybrid current drive (LHCD) is scaled by coupling the saturation process of parametric decay instability induced by LH waves in the scrape off layer (SOL) plasma to the propagation of waves. It is shown that the density limit of LHCD satisfies $n_\text{lim}\propto L_y^{2/3} P_0^{-2/3}\omega_0^{2}B_0^{2/3}T_e$, which is consistent with results of simulations and previous experiments. Both theoretical analysis and simulation results indicate that the density limit is far from being reached for ITER baseline profile. Therefore, the density limit phenomena will not prevent LHCD from being a promising method of driving plasma current at ITER and future tokamaks.

Parametric point clouds are sampled from CAD shapes, and have become increasingly prevalent in industrial manufacturing. However, most existing point cloud learning methods focus on the geometric features, such as developing efficient convolution operations, overlooking the important attribute of constraints inherent in CAD shapes, which limits these methods' ability to comprehend CAD shapes fully. To address this issue, we analyzed the effect of constraints, and proposed its deep learning-friendly representation, after that, the Constraint Feature Learning Network (CstNet) was developed to extract and leverage constraints. Our CstNet includes two stages. Stage 1 extracts constraints from B-Rep data or point cloud. Stage 2 leverages coordinates and constraints to enhance the comprehension of CAD shapes. Additionally, we built up the Parametric 20,000 Multi-modal Dataset for the scarcity of labeled B-Rep datasets. Experiments demonstrate that our CstNet achieved state-of-the-art performance on both public and proposed CAD shape datasets. To the best of our knowledge, CstNet is the first constraint-based learning method tailored for CAD shape analysis.

Drawing freehand sketches of mechanical components on multimedia devices for AI-based engineering modeling has become a new trend. However, its development is being impeded because existing works cannot produce suitable sketches for data-driven research. These works either generate sketches lacking a freehand style or utilize generative models not originally designed for this task resulting in poor effectiveness. To address this issue, we design a two-stage generative framework mimicking the human sketching behavior pattern, called MSFormer, which is the first time to produce humanoid freehand sketches tailored for mechanical components. The first stage employs Open CASCADE technology to obtain multi-view contour sketches from mechanical components, filtering perturbing signals for the ensuing generation process. Meanwhile, we design a view selector to simulate viewpoint selection tasks during human sketching for picking out information-rich sketches. The second stage translates contour sketches into freehand sketches by a transformer-based generator. To retain essential modeling features as much as possible and rationalize stroke distribution, we introduce a novel edge-constraint stroke initialization. Furthermore, we utilize a CLIP vision encoder and a new loss function incorporating the Hausdorff distance to enhance the generalizability and robustness of the model. Extensive experiments demonstrate that our approach achieves state-of-the-art performance for generating freehand sketches in the mechanical domain. Project page: https://mcfreeskegen.github.io ., Comment: Published at ACM Multimedia (ACM MM) 2024

In recent years, the rapid increase in scientific papers has overwhelmed traditional review mechanisms, resulting in varying quality of publications. Although existing methods have explored the capabilities of Large Language Models (LLMs) for automated scientific reviewing, their generated contents are often generic or partial. To address the issues above, we introduce an automated paper reviewing framework SEA. It comprises of three modules: Standardization, Evaluation, and Analysis, which are represented by models SEA-S, SEA-E, and SEA-A, respectively. Initially, SEA-S distills data standardization capabilities of GPT-4 for integrating multiple reviews for a paper. Then, SEA-E utilizes standardized data for fine-tuning, enabling it to generate constructive reviews. Finally, SEA-A introduces a new evaluation metric called mismatch score to assess the consistency between paper contents and reviews. Moreover, we design a self-correction strategy to enhance the consistency. Extensive experimental results on datasets collected from eight venues show that SEA can generate valuable insights for authors to improve their papers., Comment: Accepted by EMNLP 2024

Most existing robotic datasets capture static scene data and thus are limited in evaluating robots' dynamic performance. To address this, we present a mobile robot oriented large-scale indoor dataset, denoted as THUD (Tsinghua University Dynamic) robotic dataset, for training and evaluating their dynamic scene understanding algorithms. Specifically, the THUD dataset construction is first detailed, including organization, acquisition, and annotation methods. It comprises both real-world and synthetic data, collected with a real robot platform and a physical simulation platform, respectively. Our current dataset includes 13 larges-scale dynamic scenarios, 90K image frames, 20M 2D/3D bounding boxes of static and dynamic objects, camera poses, and IMU. The dataset is still continuously expanding. Then, the performance of mainstream indoor scene understanding tasks, e.g. 3D object detection, semantic segmentation, and robot relocalization, is evaluated on our THUD dataset. These experiments reveal serious challenges for some robot scene understanding tasks in dynamic scenes. By sharing this dataset, we aim to foster and iterate new mobile robot algorithms quickly for robot actual working dynamic environment, i.e. complex crowded dynamic scenes., Comment: This version has been accepted by ICRA2024 and the dataset has been published, where the link can be found in the paper

In the present paper, we study the properties of $\phi$-meson longitudinal leading-twist light-cone distribution amplitude $\phi_{2;{\phi}}^{\|}(x,\mu)$ by starting from a light-cone harmonic oscillator model for its wavefunction. To fix the input parameters, we derive the first ten $\xi$-moments of $\phi_{2;{\phi}}^{\|}(x,\mu)$ by using the QCD sum rules approach under the background field theory. The shape of $\phi_{2;{\phi}}^{\|}(x,\mu=2~{\rm GeV})$ tends to be a single-peak behavior, which is consistent with the latest Lattice QCD result. As an application, we derive the $D^+_s \to \phi$ transition form factors (TFFs) by using the light-cone sum rules approach. At the large recoil point, we obtain $A_1(0) = 0.512_{-0.020}^{+0.030}$, $A_2(0) = 0.402_{-0.067}^{+0.078}$, $A_0(0) = 0.596_{-0.020}^{+0.025}$ and $V(0) = 0.882_{-0.036}^{+0.040}$. As for the two typical ratios $\gamma_V$ and $\gamma_2$, we obtain $\gamma_V = 1.723_{-0.021}^{+0.023}$ and $\gamma_2 = 0.785_{-0.104}^{+0.100}$. After extrapolating those TFFs to the physically allowable region, we then obtain the transverse, longitudinal and total decay widths for semi-leptonic decay $D^+_s\to\phi\ell^+\nu_{\ell}$. Then the branching fractions are ${\cal B}(D^+_s\to \phi e^+\nu_e) = (2.367_{-0.132}^{+0.256})\times 10^{-3}$ and ${\cal B}(D^+_s\to \phi \mu^+\nu_{\mu}) = (2.349_{-0.132}^{+0.255})\times 10^{-3}$, which show good agreement with the data issued by the BESIII, the CLEO, and the BABAR Collaborations. We finally calculate $D^+_s\to\phi\ell^+ \nu_\ell$ polarization and asymmetry parameters., Comment: 32 pages, 7 figures, comments welcome

Despite the success in 6D pose estimation in bin-picking scenarios, existing methods still struggle to produce accurate prediction results for symmetry objects and real world scenarios. The primary bottlenecks include 1) the ambiguity keypoints caused by object symmetries; 2) the domain gap between real and synthetic data. To circumvent these problem, we propose a new 6D pose estimation network with symmetric-aware keypoint prediction and self-training domain adaptation (SD-Net). SD-Net builds on pointwise keypoint regression and deep hough voting to perform reliable detection keypoint under clutter and occlusion. Specifically, at the keypoint prediction stage, we designe a robust 3D keypoints selection strategy considering the symmetry class of objects and equivalent keypoints, which facilitate locating 3D keypoints even in highly occluded scenes. Additionally, we build an effective filtering algorithm on predicted keypoint to dynamically eliminate multiple ambiguity and outlier keypoint candidates. At the domain adaptation stage, we propose the self-training framework using a student-teacher training scheme. To carefully distinguish reliable predictions, we harnesses a tailored heuristics for 3D geometry pseudo labelling based on semi-chamfer distance. On public Sil'eane dataset, SD-Net achieves state-of-the-art results, obtaining an average precision of 96%. Testing learning and generalization abilities on public Parametric datasets, SD-Net is 8% higher than the state-of-the-art method. The code is available at https://github.com/dingthuang/SD-Net.

Existing Object Pose Estimation (OPE) methods for stacked scenarios are not robust to changes in object scale. This paper proposes a new 6DoF OPE network (NormNet) for different scale objects in stacked scenarios. Specifically, each object's scale is first learned with point-wise regression. Then, all objects in the stacked scenario are normalized into the same scale through semantic segmentation and affine transformation. Finally, they are fed into a shared pose estimator to recover their 6D poses. In addition, we introduce a new Sim-to-Real transfer pipeline, combining style transfer and domain randomization. This improves the NormNet's performance on real data even if we only train it on synthetic data. Extensive experiments demonstrate that the proposed method achieves state-of-the-art performance on public benchmarks and the MultiScale dataset we constructed. The real-world experiments show that our method can robustly estimate the 6D pose of objects at different scales.

Transformers have achieved significant success in medical image segmentation, owing to its capability to capture long-range dependencies. Previous works incorporate convolutional layers into the encoder module of transformers, thereby enhancing their ability to learn local relationships among pixels. However, transformers may suffer from limited generalization capabilities and reduced robustness, attributed to the insufficient spatial recovery ability of their decoders. To address this issue, A convolution sparse vector coding based decoder is proposed , namely CAScaded multi-layer Convolutional Sparse vector Coding DEcoder (CASCSCDE), which represents features extracted by the encoder using sparse vectors. To prove the effectiveness of our CASCSCDE, The widely-used TransUNet model is chosen for the demonstration purpose, and the CASCSCDE is incorporated with TransUNet to establish the TransCASCSCDE architecture. Our experiments demonstrate that TransUNet with CASCSCDE significantly enhances performance on the Synapse benchmark, obtaining up to 3.15\% and 1.16\% improvements in DICE and mIoU scores, respectively. CASCSCDE opens new ways for constructing decoders based on convolutional sparse vector coding., Comment: 8 pages, 1 figures

Recent works have shown that Large Language Models (LLMs) can facilitate the grounding of instructions for robotic task planning. Despite this progress, most existing works have primarily focused on utilizing raw images to aid LLMs in understanding environmental information. However, this approach not only limits the scope of observation but also typically necessitates extensive multimodal data collection and large-scale models. In this paper, we propose a novel approach called Graph-based Robotic Instruction Decomposer (GRID), which leverages scene graphs instead of images to perceive global scene information and iteratively plan subtasks for a given instruction. Our method encodes object attributes and relationships in graphs through an LLM and Graph Attention Networks, integrating instruction features to predict subtasks consisting of pre-defined robot actions and target objects in the scene graph. This strategy enables robots to acquire semantic knowledge widely observed in the environment from the scene graph. To train and evaluate GRID, we establish a dataset construction pipeline to generate synthetic datasets for graph-based robotic task planning. Experiments have shown that our method outperforms GPT-4 by over 25.4% in subtask accuracy and 43.6% in task accuracy. Moreover, our method achieves a real-time speed of 0.11s per inference. Experiments conducted on datasets of unseen scenes and scenes with varying numbers of objects demonstrate that the task accuracy of GRID declined by at most 3.8%, showcasing its robust cross-scene generalization ability. We validate our method in both physical simulation and the real world. More details can be found on the project page https://jackyzengl.github.io/GRID.github.io/., Comment: 8 pages, 10 figures

The $\eta^{(\prime)}$-mesons in the quark-flavor basis are mixtures of two mesonic states $|\eta_{q}\rangle=|\bar u u+\bar d d\rangle/\sqrt 2$ and $|\eta_{s}\rangle=|\bar s s\rangle$. In the previous work, we have made a detailed study on the $\eta_{s}$ leading-twist distribution amplitude. As a sequential work, in the present paper, we fix the $\eta_q$ leading-twist distribution amplitude by using the light-cone harmonic oscillator model for its wave function and by using the QCD sum rules within the QCD background field to calculate its moments. The input parameters of $\eta_q$ leading-twist distribution amplitude $\phi_{2;\eta_q}$ at an initial scale $\mu_0\sim 1$ GeV are then fixed by using those moments. The sum rules for the $0_{\rm th}$-order moment can also be used to fix the magnitude of $\eta_q$ decay constant, which gives $f_{\eta_q}=0.141\pm0.005$ GeV. As an application of the present derived $\phi_{2;\eta_q}$, we calculate the transition form factors $B(D)^+ \to\eta^{(\prime)}$ by using the QCD light-cone sum rules up to twist-4 accuracy and by including the next-to-leading order QCD corrections to the twist-2 part, and then fix the related CKM matrix element and the decay width for the semi-leptonic decays $B(D)^+ \to\eta^{(\prime)}\ell^+ \nu_\ell$., Comment: 16 pages, 8 figures, Comments welcome

Grasping an object when it is in an ungraspable pose is a challenging task, such as books or other large flat objects placed horizontally on a table. Inspired by human manipulation, we address this problem by pushing the object to the edge of the table and then grasping it from the hanging part. In this paper, we develop a model-free Deep Reinforcement Learning framework to synergize pushing and grasping actions. We first pre-train a Variational Autoencoder to extract high-dimensional features of input scenario images. One Proximal Policy Optimization algorithm with the common reward and sharing layers of Actor-Critic is employed to learn both pushing and grasping actions with high data efficiency. Experiments show that our one network policy can converge 2.5 times faster than the policy using two parallel networks. Moreover, the experiments on unseen objects show that our policy can generalize to the challenging case of objects with curved surfaces and off-center irregularly shaped objects. Lastly, our policy can be transferred to a real robot without fine-tuning by using CycleGAN for domain adaption and outperforms the push-to-wall baseline., Comment: Accepted by ICRA 2023; Project page https://haozhang990127.github.io/PaG/

Most existing image inpainting algorithms are based on a single view, struggling with large holes or the holes containing complicated scenes. Some reference-guided algorithms fill the hole by referring to another viewpoint image and use 2D image alignment. Due to the camera imaging process, simple 2D transformation is difficult to achieve a satisfactory result. In this paper, we propose 3DFill, a simple and efficient method for reference-guided image inpainting. Given a target image with arbitrary hole regions and a reference image from another viewpoint, the 3DFill first aligns the two images by a two-stage method: 3D projection + 2D transformation, which has better results than 2D image alignment. The 3D projection is an overall alignment between images and the 2D transformation is a local alignment focused on the hole region. The entire process of image alignment is self-supervised. We then fill the hole in the target image with the contents of the aligned image. Finally, we use a conditional generation network to refine the filled image to obtain the inpainting result. 3DFill achieves state-of-the-art performance on image inpainting across a variety of wide view shifts and has a faster inference speed than other inpainting models.

Some robust point cloud registration approaches with controllable pose refinement magnitude, such as ICP and its variants, are commonly used to improve 6D pose estimation accuracy. However, the effectiveness of these methods gradually diminishes with the advancement of deep learning techniques and the enhancement of initial pose accuracy, primarily due to their lack of specific design for pose refinement. In this paper, we propose Point Cloud Completion and Keypoint Refinement with Fusion Data (PCKRF), a new pose refinement pipeline for 6D pose estimation. The pipeline consists of two steps. First, it completes the input point clouds via a novel pose-sensitive point completion network. The network uses both local and global features with pose information during point completion. Then, it registers the completed object point cloud with the corresponding target point cloud by our proposed Color supported Iterative KeyPoint (CIKP) method. The CIKP method introduces color information into registration and registers a point cloud around each keypoint to increase stability. The PCKRF pipeline can be integrated with existing popular 6D pose estimation methods, such as the full flow bidirectional fusion network, to further improve their pose estimation accuracy. Experiments demonstrate that our method exhibits superior stability compared to existing approaches when optimizing initial poses with relatively high precision. Notably, the results indicate that our method effectively complements most existing pose estimation techniques, leading to improved performance in most cases. Furthermore, our method achieves promising results even in challenging scenarios involving textureless and symmetrical objects. Our source code is available at https://github.com/zhanhz/KRF.

Pseudo-label-based semi-supervised learning (SSL) has achieved great success on raw data utilization. However, its training procedure suffers from confirmation bias due to the noise contained in self-generated artificial labels. Moreover, the model's judgment becomes noisier in real-world applications with extensive out-of-distribution data. To address this issue, we propose a general method named Class-aware Contrastive Semi-Supervised Learning (CCSSL), which is a drop-in helper to improve the pseudo-label quality and enhance the model's robustness in the real-world setting. Rather than treating real-world data as a union set, our method separately handles reliable in-distribution data with class-wise clustering for blending into downstream tasks and noisy out-of-distribution data with image-wise contrastive for better generalization. Furthermore, by applying target re-weighting, we successfully emphasize clean label learning and simultaneously reduce noisy label learning. Despite its simplicity, our proposed CCSSL has significant performance improvements over the state-of-the-art SSL methods on the standard datasets CIFAR100 and STL10. On the real-world dataset Semi-iNat 2021, we improve FixMatch by 9.80% and CoMatch by 3.18%. Code is available https://github.com/TencentYoutuResearch/Classification-SemiCLS., Comment: cvpr2022 accepted, half more page for adding rebuttal Infos

In the paper, we study the $B_s\to D_s$ transition form factors by using the light-cone sum rules within the framework of heavy quark effective field theory. We adopt a chiral current correlation function to do the calculation, the resultant transition form factors $f_+^{B_s\to D_s}(q^2)$ and $f_0^{B_s\to D_s}(q^2)$ are dominated by the contribution of $D_s$-meson leading-twist distribution amplitude, while the contributions from less certain $D_s$-meson twist-3 distribution amplitudes are greatly suppressed. At the largest recoil point, we obtain $f_{+,0}^{B_s \to D_s}(0)=0.533^{+0.112}_{-0.094}$. By further extrapolating the transition form factors into all the physically allowable $q^2$ region with the help of the $z$-series parametrization approach, we calculate the branching fractions $\mathcal{B}(B_s \to D_s \ell^\prime \nu_{\ell^\prime})$ with $(\ell^\prime= e,\mu)$ and $\mathcal{B}(B_s \to D_s \tau \nu_\tau)$, which gives $\mathcal{R}(D_s)=0.334\pm 0.017$., Comment: 9 pages, 7 figures

The China Fusion Engineering Test Reactor (CFETR) and the Huazhong Field Reversed Configuration (HFRC), currently both under intensive physical and engineering designs in China, are the two major projects representative of the low-density steady-state and high-density pulsed pathways to fusion. One of the primary tasks of the physics designs for both CFETR and HFRC is the assessment and analysis of the magnetohydrodynamic (MHD) stability of the proposed design schemes. Comprehensive efforts on the assessment of MHD stability of CFETR and HFRC baseline scenarios have led to preliminary progresses that may further benefit engineering designs., Comment: 48 pages, 26 figures

In this paper, we calculate the $B_c\to J/\psi$ helicity form factors (HFFs) up to twist-4 accuracy by using the light-cone sum rules (LCSR) approach. After extrapolating those HFFs to the physically allowable $q^2$ region, we investigate the $B^+_c$-meson two-body decays and semi-leptonic decays $B_c^+ \to J/\psi+(P, V, \ell^+ \nu_\ell)$ with $P/V$ stands for light pseudoscalar/vector meson, respectively. The branching fractions can be derived by using the CKM matrix element and the $B_c$ lifetime from the Particle Data Group, and we obtain ${\cal B}(B_c^+ \to J/\psi \pi^+)=(0.136^{+0.002}_{-0.002})\%$, ${\cal B}(B_c^+ \to J/\psi K^+)=(0.010^{+0.000}_{-0.000})\%$, ${\cal B}(B_c^+ \to J/\psi \rho^+) =(0.768^{+0.029}_{-0.033})\%$, ${\cal B}(B_c^+ \to J/\psi K^{\ast +})=(0.043^{+0.001}_{-0.001})\%$, ${\cal B}(B_c^+ \to J/\psi \mu^+\nu_\mu)=(2.802^{+0.526}_{-0.675})\%$ and ${\cal B}(B_c^+ \to J/\psi \tau^+\nu_\tau)=(0.559^{+0.131}_{-0.170})\%$. We then obtain ${\cal R}_{\pi^+/\mu^+\nu_\mu} = 0.048^{+ 0.009}_{-0.012}$ and ${\cal R}_{K^+ / \pi^+} = 0.075^{+0.005}_{-0.005}$, which agree with the LHCb measured value within $1\sigma$-error. We also obtain ${\cal R}_{J/\psi}=0.199^{+ 0.060}_{-0.077}$, which like other theoretical predictions, is consistent with the LHCb measured value within $2\sigma$-error. Those imply that the HFFs under the LCSR approach are also applicable to the $B^+_c$ meson two-body decays and semi-leptonic decays $B_c^+ \to J/\psi+(P, V, \ell^+ \nu_\ell)$, and the HFFs obtained by using LCSR in a new way implies that there may be new physics in the $B_c\to J/\psi \ell^+ \nu_\ell$ semi-leptonic decays., Comment: 10 pages, 4 figures, published version

In this paper, we make a detailed discussion on the $\eta$ and $\eta'$-meson leading-twist light-cone distribution amplitude $\phi_{2;\eta^{(\prime)}}(u,\mu)$ by using QCD sum rules approach under the background field theory. Taking both the non-perturbative condensates up to dimension-six and NLO QCD corrections to the perturbative part, its first three moments $\langle\xi^n_{2;\eta^{(\prime)}}\rangle|_{\mu_0} $ with $n = (2,4,6)$ at initial scale $\mu_0 = 1$ GeV can be determined. e.g. $\langle\xi_{2;\eta}^2\rangle|_{\mu_0} =0.231_{-0.013}^{+0.010}$, $\langle\xi_{2;\eta}^4 \rangle|_{\mu_0} =0.109_{-0.007}^{+0.007}$, and $\langle\xi_{2;\eta}^6 \rangle|_{\mu_0} =0.066_{-0.006}^{+0.006}$ for $\eta$-meson, $\langle\xi_{2;\eta'}^2\rangle|_{\mu_0} =0.211_{-0.017}^{+0.015}$, $\langle\xi_{2;\eta'}^4 \rangle|_{\mu_0} =0.093_{-0.009}^{+0.009}$, and $\langle\xi_{2;\eta'}^6 \rangle|_{\mu_0} =0.054_{-0.008}^{+0.008}$ for $\eta'$-meson. Next, we calculate $D_s\to\eta^{(\prime)}$ TFFs $f^{\eta^{(\prime)}}_+(q^2)$ within QCD light-cone sum rules approach up to NLO level. The values at large recoil region are $f^{\eta}_+(0) = 0.476_{-0.036}^{+0.040}$ and $f^{\eta'}_+(0) = 0.544_{-0.042}^{+0.046}$. After extrapolating TFFs to the allowable physical regions within the series expansion, we obtain the branching fractions of the semi-leptonic decay, i.e. $D_s^+\to\eta^{(\prime)}\ell^+ \nu_\ell$, i.e. ${\cal B}(D_s^+\to\eta^{(\prime)} e^+\nu_e)=2.346_{-0.331}^{+0.418}(0.792_{-0.118}^{+0.141})\times10^{-2}$ and ${\cal B}(D_s^+\to\eta^{(\prime)} \mu^+\nu_\mu)=2.320_{-0.327}^{+0.413}(0.773_{-0.115}^{+0.138})\times10^{-2}$ for $\ell = (e, \mu)$ channels respectively. And in addition to that, the mixing angle for $\eta-\eta'$ with $\varphi$ and ratio for the different decay channels ${\cal R}_{\eta'/\eta}^\ell$ are given, which show good agreement with the recent BESIII measurements., Comment: 52 pages, 11 figures, version to be published in EPJC

We make a detailed study on the typical production channel of double charmoniums, $e^+e^-\to J/\psi+\eta_c$, at the center-of-mass collision energy $\sqrt{s}=10.58$ GeV. The key component of the process is the form factor $F_{\rm VP}(q^2)$, which has been calculated within the QCD light-cone sum rules (LCSR). To improve the accuracy of the derived LCSR, we keep the $J/\psi$ light-cone distribution amplitude up to twist-4 accuracy. Total cross sections for $e^+e^-\to J/\psi+\eta_c$ at three typical factorization scales are $\sigma|_{\mu_s} = 22.53^{+3.46}_{-3.49}~{\rm fb}$, $\sigma|_{\mu_k} = 21.98^{+3.35}_{-3.38}~{\rm fb}$ and $\sigma|_{\mu_0} = 21.74^{+3.29}_{-3.33}~{\rm fb}$, respectively. The factorization scale dependence is small, and those predictions are consistent with the BABAR and Belle measurements within errors., Comment: 8 pages, 3 figures, version to be published in Physical Review D