arXiv subjects

Yang Gao

Publications and source records attributed to Yang Gao.

At least 415 records · Page 23Linked to original sources

arXiv2024

MQE: Unleashing the Power of Interaction with Multi-agent Quadruped Environment

The advent of deep reinforcement learning (DRL) has significantly advanced the field of robotics, particularly in the control and coordination of quadruped robots. However, the complexity of real-world tasks often necessitates the deployment of multi-robot systems capable of sophisticated interaction and collaboration. To address this need, we introduce the Multi-agent Quadruped Environment (MQE), a novel platform designed to facilitate the development and evaluation of multi-agent reinforcement learning (MARL) algorithms in realistic and dynamic scenarios. MQE emphasizes complex interactions between robots and objects, hierarchical policy structures, and challenging evaluation scenarios that reflect real-world applications. We present a series of collaborative and competitive tasks within MQE, ranging from simple coordination to complex adversarial interactions, and benchmark state-of-the-art MARL algorithms. Our findings indicate that hierarchical reinforcement learning can simplify task learning, but also highlight the need for advanced algorithms capable of handling the intricate dynamics of multi-agent interactions. MQE serves as a stepping stone towards bridging the gap between simulation and practical deployment, offering a rich environment for future research in multi-agent systems and robot learning. For open-sourced code and more details of MQE, please refer to https://ziyanx02.github.io/multiagent-quadruped-environment/ .

cs.RO↗

arXiv2024

Precise measurement of the $e^+e^-\to D_s^+D_s^-$ cross sections at center-of-mass energies from threshold to 4.95 GeV

Using the $e^+e^-$ collision data collected with the BESIII detector operating at the BEPCII collider, at center-of-mass energies from the threshold to $4.95$~GeV, we present precise measurements of the cross sections for the process $e^+e^-\to D_s^+D_s^-$ using a single tag method. The resulting cross section lineshape exhibits several new structures, thereby offering an input for coupled channel analysis and model tests, which are critical to understand vector charmonium-like states with masses between 4 and 5~GeV.

hep-ex↗

arXiv2024

QSMDiff: Unsupervised 3D Diffusion Models for Quantitative Susceptibility Mapping

Quantitative Susceptibility Mapping (QSM) dipole inversion is an ill-posed inverse problem for quantifying magnetic susceptibility distributions from MRI tissue phases. While supervised deep learning methods have shown success in specific QSM tasks, their generalizability across different acquisition scenarios remains constrained. Recent developments in diffusion models have demonstrated potential for solving 2D medical imaging inverse problems. However, their application to 3D modalities, such as QSM, remains challenging due to high computational demands. In this work, we developed a 3D image patch-based diffusion model, namely QSMDiff, for robust QSM reconstruction across different scan parameters, alongside simultaneous super-resolution and image-denoising tasks. QSMDiff adopts unsupervised 3D image patch training and full-size measurement guidance during inference for controlled image generation. Evaluation on simulated and in-vivo human brains, using gradient-echo and echo-planar imaging sequences across different acquisition parameters, demonstrates superior performance. The method proposed in QSMDiff also holds promise for impacting other 3D medical imaging applications beyond QSM.

eess.IV↗

arXiv2024

Uni-O4: Unifying Online and Offline Deep Reinforcement Learning with Multi-Step On-Policy Optimization

Combining offline and online reinforcement learning (RL) is crucial for efficient and safe learning. However, previous approaches treat offline and online learning as separate procedures, resulting in redundant designs and limited performance. We ask: Can we achieve straightforward yet effective offline and online learning without introducing extra conservatism or regularization? In this study, we propose Uni-o4, which utilizes an on-policy objective for both offline and online learning. Owning to the alignment of objectives in two phases, the RL agent can transfer between offline and online learning seamlessly. This property enhances the flexibility of the learning paradigm, allowing for arbitrary combinations of pretraining, fine-tuning, offline, and online learning. In the offline phase, specifically, Uni-o4 leverages diverse ensemble policies to address the mismatch issues between the estimated behavior policy and the offline dataset. Through a simple offline policy evaluation (OPE) approach, Uni-o4 can achieve multi-step policy improvement safely. We demonstrate that by employing the method above, the fusion of these two paradigms can yield superior offline initialization as well as stable and rapid online fine-tuning capabilities. Through real-world robot tasks, we highlight the benefits of this paradigm for rapid deployment in challenging, previously unseen real-world environments. Additionally, through comprehensive evaluations using numerous simulated benchmarks, we substantiate that our method achieves state-of-the-art performance in both offline and offline-to-online fine-tuning learning. Our website: https://lei-kun.github.io/uni-o4/ .

cs.LG↗

arXiv2024

AMSP: Reducing Communication Overhead of ZeRO for Efficient LLM Training

Training large language models (LLMs) encounters challenges in GPU memory consumption due to the high memory requirements of model states. The widely used Zero Redundancy Optimizer (ZeRO) addresses this issue through strategic sharding but introduces communication challenges at scale. To tackle this problem, we propose AMSP, a system designed to optimize ZeRO for scalable LLM training. AMSP incorporates three flexible sharding strategies: Full-Replica, Full-Sharding, and Partial-Sharding, and allows each component within the model states (Parameters, Gradients, Optimizer States) to independently choose a sharding strategy as well as the device mesh. We conduct a thorough analysis of communication costs, formulating an optimization problem to discover the optimal sharding strategy. Additionally, AMSP optimizes distributed LLM training by efficiently overlapping communication with computation. Evaluations demonstrate up to 52\% Model FLOPs Utilization (MFU) when training the LLaMA-based model on 1024 GPUs, resulting in a 1.56 times improvement in training throughput compared to newly proposed systems like MiCS and ZeRO++.

cs.DC↗

arXiv2024

SpaceOctopus: An Octopus-inspired Motion Planning Framework for Multi-arm Space Robot

Space robots have played a critical role in autonomous maintenance and space junk removal. Multi-arm space robots can efficiently complete the target capture and base reorientation tasks due to their flexibility and the collaborative capabilities between the arms. However, the complex coupling properties arising from both the multiple arms and the free-floating base present challenges to the motion planning problems of multi-arm space robots. We observe that the octopus elegantly achieves similar goals when grabbing prey and escaping from danger. Inspired by the distributed control of octopuses' limbs, we develop a multi-level decentralized motion planning framework to manage the movement of different arms of space robots. This motion planning framework integrates naturally with the multi-agent reinforcement learning (MARL) paradigm. The results indicate that our method outperforms the previous method (centralized training). Leveraging the flexibility of the decentralized framework, we reassemble policies trained for different tasks, enabling the space robot to complete trajectory planning tasks while adjusting the base attitude without further learning. Furthermore, our experiments confirm the superior robustness of our method in the face of external disturbances, changing base masses, and even the failure of one arm.

cs.RO↗

arXiv2024

CoPa: General Robotic Manipulation through Spatial Constraints of Parts with Foundation Models

Foundation models pre-trained on web-scale data are shown to encapsulate extensive world knowledge beneficial for robotic manipulation in the form of task planning. However, the actual physical implementation of these plans often relies on task-specific learning methods, which require significant data collection and struggle with generalizability. In this work, we introduce Robotic Manipulation through Spatial Constraints of Parts (CoPa), a novel framework that leverages the common sense knowledge embedded within foundation models to generate a sequence of 6-DoF end-effector poses for open-world robotic manipulation. Specifically, we decompose the manipulation process into two phases: task-oriented grasping and task-aware motion planning. In the task-oriented grasping phase, we employ foundation vision-language models (VLMs) to select the object's grasping part through a novel coarse-to-fine grounding mechanism. During the task-aware motion planning phase, VLMs are utilized again to identify the spatial geometry constraints of task-relevant object parts, which are then used to derive post-grasp poses. We also demonstrate how CoPa can be seamlessly integrated with existing robotic planning algorithms to accomplish complex, long-horizon tasks. Our comprehensive real-world experiments show that CoPa possesses a fine-grained physical understanding of scenes, capable of handling open-set instructions and objects with minimal prompt engineering and without additional training. Project page: https://copa-2024.github.io/

cs.RO↗

arXiv2024

Observation of the decay $h_{c}\to3(π^{+}π^{-})π^{0}$

Based on $(2712.4\pm14.1)\times10^{6}$ $ψ(3686)$ events collected with the BESIII detector, we study the decays $h_{c}\to3(π^{+}π^{-})π^{0}$, $h_{c}\to2(π^{+}π^{-})ω$, $h_{c}\to2(π^{+}π^{-})π^{0}η$, $h_{c}\to2(π^{+}π^{-})η$, and $h_{c}\to p\bar{p}$ via $ψ(3686)\toπ^{0}h_{c}$. The decay channel $h_{c}\to3(π^{+}π^{-})π^{0}$ is observed for the first time, and its branching fraction is determined to be $\left( {9.28\pm 1.14 \pm 0.77} \right) \times {10^{ - 3}}$, where the first uncertainty is statistical and the second is systematic. In addition, first evidence is found for the modes $h_{c} \to 2(π^{+}π^{-})π^{0}η$ and $h_{c}\to2(π^{+}π^{-})ω$ with significances of 4.8$σ$ and 4.7$σ$, and their branching fractions are determined to be $(7.55\pm1.51\pm0.77)\times10^{-3}$ and $\left( {4.00 \pm 0.86 \pm 0.35}\right) \times {10^{ - 3}}$, respectively. No significant signals of $h_c\to 2(π^+π^-)η$ and $h_{c}\to p\bar{p}$ are observed, and the upper limits of the branching fractions of these decays are determined to be $<6.19\times10^{-4}$ and $<4.40\times10^{-5}$ at the 90% confidence level, respectively.

hep-ex↗

arXiv2024

Determination of the $Σ^{+}$ Timelike Electromagnetic Form Factors

Based on data samples collected with the BESIII detector at the BEPCII collider, the process $e^{+}e^{-} \to Σ^{+}\barΣ^{-}$ is studied at center-of-mass energies $\sqrt{s}$ = 2.3960, 2.6454, and 2.9000 GeV. Using a fully differential angular description of the final state particles, both the relative magnitude and phase information of the $Σ^{+}$ electromagnetic form factors in the timelike region are extracted. The relative phase between the electric and magnetic form factors is determined to be $\sinΔΦ$ = -0.67~$\pm$~0.29~(stat)~$\pm$~0.18~(syst) at $\sqrt{s}$ = 2.3960 GeV, $ΔΦ$ = 55$^{\circ}$~$\pm$~19$^{\circ}$~(stat) $\pm$~14$^{\circ}$~(syst) at $\sqrt{s}$ = 2.6454 GeV, and 78$^{\circ}$~$\pm$~22$^{\circ}$~(stat) $\pm$~9$^{\circ}$~(syst) at $\sqrt{s}$ = 2.9000 GeV. For the first time, the phase of the hyperon electromagnetic form factors is explored in a wide range of four-momentum transfer. The evolution of the phase along with four-momentum transfer is an important input for understanding its asymptotic behavior and the dynamics of baryons.

hep-ex↗

arXiv2024

Observation of $ψ(3686)\to 3ϕ$

Using $(2.712\pm0.014)\times 10^9$ $ψ(3686)$ events collected by the BESIII detector operating at the BEPCII collider, we report the first observation of $ψ(3686)\to 3ϕ$ decay with a significance larger than 10$σ$. The branching fraction of this decay is determined to be $(1.46\pm0.05\pm0.17)\times10^{-5}$, where the first uncertainty is statistical and the second is systematic. No significant structure is observed in the $ϕϕ$ invariant mass spectra.

hep-ex↗

arXiv2024

Updated measurement of the branching fraction of $D_s^+\to τ^+ν_τ$ via $τ^+\toπ^+\barν_τ$

This paper reports the study of $D_s^+\to τ^+ν$ via $τ^+\toπ^+\barν_τ$ using a boosted decision tree method, with $7.33$ fb$^{-1}$ of $e^{+}e^{-}$ collision data collected by the BESIII detector at center-of-mass energies between $4.128$ and $4.226$~GeV. The branching fraction of $D_s^+\to τ^+ν_τ$ is determined to be $(5.44\pm0.17_{\rm stat}\pm0.13_{\rm syst})\%$. The product of the $D_s^+$ decay constant $f_{D_s^+}$ and the CKM matrix element $|V_{cs}|$ is $f_{D_s^+}|V_{cs}| = (248.3\pm3.9_{\rm stat}\pm3.1_{\rm syst}\pm1.0_{\rm input})~\mathrm{MeV}$. Combining with the $|V_{cs}|$ value obtained from the Standard Model global fit or the $f_{D_s^+}$ from the lattice quantum chromodynamics, we determine $|V_{cs}| = 0.993\pm0.015_{\rm stat}\pm0.012_{\rm syst}\pm0.004_{\rm input}$ and $f_{D_s^+} = (255.0\pm4.0_{\rm stat}\pm3.2_{\rm syst}\pm1.0_{\rm input})~\text{MeV}$. The first uncertainty is statistical, the second one is systematic and the third one is due to the input parameters, mainly the lifetime of $D_s^+$. All results obtained in this work supersede the BESIII previous results based on 6.32 fb$^{-1}$ of $e^+e^-$ collision data taken at center-of-mass energies between $4.178$ and $4.226$~GeV.

hep-ex↗

arXiv2024

Revisiting Disentanglement in Downstream Tasks: A Study on Its Necessity for Abstract Visual Reasoning

In representation learning, a disentangled representation is highly desirable as it encodes generative factors of data in a separable and compact pattern. Researchers have advocated leveraging disentangled representations to complete downstream tasks with encouraging empirical evidence. This paper further investigates the necessity of disentangled representation in downstream applications. Specifically, we show that dimension-wise disentangled representations are unnecessary on a fundamental downstream task, abstract visual reasoning. We provide extensive empirical evidence against the necessity of disentanglement, covering multiple datasets, representation learning methods, and downstream network architectures. Furthermore, our findings suggest that the informativeness of representations is a better indicator of downstream performance than disentanglement. Finally, the positive correlation between informativeness and disentanglement explains the claimed usefulness of disentangled representations in previous works. The source code is available at https://github.com/Richard-coder-Nai/disentanglement-lib-necessity.git.

cs.CV↗

arXiv2024

Can Transformers Capture Spatial Relations between Objects?

Spatial relationships between objects represent key scene information for humans to understand and interact with the world. To study the capability of current computer vision systems to recognize physically grounded spatial relations, we start by proposing precise relation definitions that permit consistently annotating a benchmark dataset. Despite the apparent simplicity of this task relative to others in the recognition literature, we observe that existing approaches perform poorly on this benchmark. We propose new approaches exploiting the long-range attention capabilities of transformers for this task, and evaluating key design principles. We identify a simple "RelatiViT" architecture and demonstrate that it outperforms all current approaches. To our knowledge, this is the first method to convincingly outperform naive baselines on spatial relation prediction in in-the-wild settings. The code and datasets are available in \url{https://sites.google.com/view/spatial-relation}.

cs.CV↗

arXiv2024

Observation of $D^+_s\to η^\prime μ^+ν_μ$, Precision Test of Lepton Flavor Universality with $D^+_s\to η^{(\prime)} \ell^+ν_\ell$, and First Measurements of $D^+_s\to η^{(\prime)}μ^+ν_μ$ Decay Dynamics

By analyzing 7.33 fb$^{-1}$ of $e^+e^-$ annihilation data collected at center-of-mass energies between 4.128 and 4.226 GeV with the BESIII detector, we report the observation of the semileptonic decay $D^+_s\to η^\prime μ^+ν_μ$, with a statistical significance larger than 10$σ$, and the measurements of the $D_s^+ \to ημ^+ν_μ$ and $D_s^+ \to η^\primeμ^+ν_μ$ decay dynamics for the first time. The branching fractions of $D_s^+ \to ημ^+ν_μ$ and $D_s^+ \to η^\primeμ^+ν_μ$ are determined to be $(2.235\pm0.051_{\rm stat}\pm0.052_{\rm syst})\%$ and $(0.801\pm0.055_{\rm stat}\pm0.028_{\rm syst})\%$, respectively, with precision improved by factors of 6.0 and 6.6 compared to the previous best measurements. Combined with the results for the decays $D_s^+ \to ηe^+ν_e$ and $D_s^+ \to η^\prime e^+ν_e$, the ratios of the decay widths are examined both inclusively and in several $\ell^+ν_\ell$ four-momentum transfer ranges. No evidence for lepton flavor universality violation is found within the current statistics. The products of the hadronic form factors $f_{+,0}^{η^{(\prime)}}(0)$ and the $c\to s$ Cabibbo-Kobayashi-Maskawa matrix element $|V_{cs}|$ are determined. The results based on the two-parameter series expansion are $f^η_{+,0}(0)|V_{cs}| = 0.452\pm0.010_{\rm stat}\pm0.007_{\rm syst}$ and $f^{η^{\prime}}_{+,0}(0)|V_{cs}| = 0.504\pm0.037_{\rm stat}\pm0.012_{\rm syst}$, which help to constrain present models on $f_{+,0}^{η^{(\prime)}}(0)$. The forward-backward asymmetries are determined to be $\langle A_{\rm FB}^η\rangle=-0.059\pm0.031_{\rm stat}\pm0.005_{\rm syst}$ and $\langle A_{\rm FB}^{η^\prime}\rangle=-0.064\pm0.079_{\rm stat}\pm0.006_{\rm syst}$ for the first time, which are consistent with the theoretical calculation.

hep-ex↗

arXiv2024

Measurement of the $e^{+}e^{-} \to K_{S}^{0} K_{L}^{0} π^{0}$ cross sections from $\sqrt{s}=$ 2.000 to 3.080 GeV

Based on $e^{+}e^{-}$ collision data collected at center-of-mass energies from 2.000 to 3.080 GeV by the BESIII detector at the BEPCII collider, a partial wave analysis is performed for the process $e^{+}e^{-}\to K_{S}^{0} K_{L}^{0} π^{0}$. The results allow the Born cross sections of the process $e^{+}e^{-}\to K_{S}^{0} K_{L}^{0} π^{0}$, as well as its subprocesses $e^{+}e^{-}\to K^{*}(892)^{0}\bar{K}^{0}$ and $K^{*}_{2}(1430)^{0}\bar{K}^{0}$ to be measured. The Born cross sections for $e^{+}e^{-}\to K_{S}^{0}K_{L}^{0}π^{0}$ are consistent with previous measurements by BaBar, but with substantially improved precision. The Born cross section lineshape of the process $e^{+}e^{-}\to K^{*}(892)^{0}\bar{K}^{0}$ is consistent with a vector meson state around 2.2 GeV with a significance of 3.2$σ$. A Breit-Wigner fit determines its mass as $M_Y=(2164.7\pm9.1\pm3.1)~{\rm{MeV}}/c^{2}$ and its width as $Γ_{Y}=(32.4\pm21.0\pm1.8)~\rm{MeV}$.

hep-ex↗

arXiv2024

First observation of $Λ_c^+\rightarrowΛK^+π^0$ and evidence of $Λ_c^+\rightarrowΛK^+π^+π^-$

We present the first observation of the singly Cabibbo-suppressed decay $Λ_c^+ \rightarrow ΛK^+π^0$ with a significance of $5.7σ$ and the first evidence of $Λ_c^+ \rightarrow ΛK^+π^+π^-$ decay with a significance of $3.1σ$, based on $e^+e^-$ annihilation data recorded by the BESIII detector at the BEPCII collider. The data correspond to an integrated luminosity of $6.4~{\rm fb^{-1}}$, in the center-of-mass energy range from $4.600~{\rm GeV}$ to $4.950~{\rm GeV}$. We determine the branching fractions of $Λ_c^+ \rightarrow ΛK^+π^0$ and $Λ_c^+ \rightarrow ΛK^+π^+π^-$ relative to their Cabibbo-favored counterparts to be $\frac{\mathcal{B}(Λ_c^+ \rightarrow ΛK^+π^0)}{\mathcal{B}(Λ_c^+ \rightarrow Λπ^+π^0)} = (2.09\pm0.39_{\mathrm{stat.}}\pm0.07_{\mathrm{syst.}}) \times 10^{-2}$ and $\frac{\mathcal{B}(Λ_c^+ \rightarrow ΛK^+π^+π^-)}{\mathcal{B}(Λ_c^+ \rightarrow Λπ^+π^+π^-)} = (1.13\pm0.41_{\mathrm{stat.}}\pm0.06_{\mathrm{syst.}}) \times 10^{-2}$, respectively. Moreover, by combining our measured result with the world average of $\mathcal{B}(Λ^+_c\to Λπ^+π^0)$, we obtain the branching fraction $\mathcal{B}(Λ_c^+ \to ΛK^+π^0) = (1.49\pm0.27_{\mathrm{stat.}}\pm0.05_{\mathrm{syst.}}\pm0.08_{\mathrm{ref.}}) \times 10^{-3}$. This result significantly departs from theoretical predictions based on quark $SU(3)$ flavor symmetry, which is underpinned by the presumption of meson pair $S$-wave amplitude dominance.

hep-ex↗

arXiv2024

Observation of the In-plane Anomalous Hall Effect induced by Octupole in Magnetization Space

The Anomalous Hall Effect (AHE) manifests as a transverse voltage proportional to magnetization in ferromagnetic materials under the application of a charge current, being an indispensable tool for probing magnetism, especially in nanoscale devices. However, the AHE primarily sensitizes to out-of-plane magnetization, thereby hindering its capacity to discern the in-plane magnetization, a characteristic prevalent in ferromagnetic films. Here we challenge this conventional understanding by demonstrating the in-plane magnetization-induced AHE in iron and nickel, two ubiquitous ferromagnets. This observation of the in-plane AHE is remarkable as it contradicts existing theories that forbid such phenomena in cubic crystal systems. We trace the origin of this unanticipated phenomenon to a hitherto unconsidered octupole of the anomalous Hall conductivity in the magnetization space, a mechanism we propose could enable the detection of in-plane AHE in a wide range of ferromagnetic materials. This work realizes the in-plane AHE in common ferromagnets by exploiting the anomalous Hall conductivity octupole, revealing a new physical origin of the AHE and promising to revolutionize the design of magnetic devices and sensors.

cond-mat.mtrl-sci↗

arXiv2024

Simulation Studies for the First Pathfinder of the CATCH Space Mission

The Chasing All Transients Constellation Hunters (CATCH) space mission is an intelligent constellation consisting of 126 micro-satellites in three types (A, B, and C), designed for X-ray observation with the objective of studying the dynamic universe. Currently, we are actively developing the first Pathfinder (CATCH-1) for the CATCH mission, specifically for type-A satellites. CATCH-1 is equipped with Micro Pore Optics (MPO) and a 4-pixel Silicon Drift Detector (SDD) array. To assess its scientific performance, including the effective area of the optical system, on-orbit background, and telescope sensitivity, we employ the Monte Carlo software Geant4 for simulation in this study. The MPO optics exhibit an effective area of $41$ cm$^2$ at the focal spot for 1 keV X-rays, while the entire telescope system achieves an effective area of $29$ cm$^2$ at 1 keV when taking into account the SDD detector's detection efficiency. The primary contribution to the background is found to be from the Cosmic X-ray Background. Assuming a 625 km orbit with an inclination of $29^\circ$, the total background for CATCH-1 is estimated to be $8.13\times10^{-2}$ counts s$^{-1}$ in the energy range of 0.5--4 keV. Based on the background within the central detector and assuming a Crab-like source spectrum, the estimated ideal sensitivity could achieve $1.9\times10^{-12}$ erg cm$^{-2}$ s$^{-1}$ for an exposure of 10$^4$ s in the energy band of 0.5--4 keV. Furthermore, after simulating the background caused by low-energy charged particles near the geomagnetic equator, we have determined that there is no need to install a magnetic deflector.

astro-ph.IM↗