arXiv Science⌕ Search

arXiv subjects

Zhipeng Cui

Publications and source records attributed to Zhipeng Cui.

4 recordsLinked to original sources

AI-Assisted Analysis of CMS Open Data: Prompt J/psi Cross-Section Measurements in pp and Peripheral PbPb Collisions at sqrt(S)= 7 TeV and sqrt(sNN) = 2.76 TeV

We investigate the use of an artificial-intelligence (AI) coding agent, supervised by human physicists, for collider physics analyses using CMS Open Data. OpenAI Codex performs dimuon event selection, two-dimensional invariant-mass and decay-length fits, Monte Carlo detector simulations, and systematic uncertainty evaluation after acquiring physics and experimental knowledge from published studies and external references, with human review at each stage. The workflow is tested on two CMS datasets, with custom full-detector simulations used to determine acceptance and reconstruction efficiencies. In proton-proton collisions at sqrt(s) = 7 TeV, prompt J/psi differential cross sections are measured versus transverse momentum and rapidity. The ratios of remeasured to published CMS cross sections have a mean of 1.000 and an RMS spread of 0.215 across 39 bins. In peripheral (50-100% centrality) PbPb collisions at sqrt(s_NN) = 2.76 TeV, the corresponding prompt J/psi fiducial cross-section ratios have a mean of 1.158 and an RMS spread of 0.225 across six intervals. We identify agent failure modes and demonstrate the importance of structured human review in ensuring reliable results.

hep-ex↗

Heavy Quarkonium Physics: An Experimental Review at Tevatron, LHC, and e+e- Colliders

We review recent experimental advances in heavy quarkonium physics at the Tevatron, e+e- colliders (Belle, BaBar, Belle II, BESIII, and LEP), and the LHC (ALICE, ATLAS, CMS, and LHCb), including Run 2 and early Run 3 results. Differential cross sections and polarization measurements of charmonium (Jpsi, Psi(2S)) and bottomonium (Upsilon(nS)) test perturbative and non-perturbative QCD within the Non-Relativistic QCD (NRQCD) framework and probe symmetry relations among spin and orbital configurations. CDF and D0 established production baselines and exposed polarization tensions with early NRQCD predictions. Complementary e+e- measurements provide precise spectroscopy, branching fractions, radiative and hadronic transitions, and observations of exotic quarkonium-like states, constraining non-perturbative dynamics. LHC measurements extend production and polarization studies across broad kinematic ranges. Spectroscopic advances include fully-heavy tetraquark candidates, exotic structures in Jpsi Jpsi and Jpsi Psi(2S) reported by LHCb, and studies of toponium (ttbar) quasi-bound states. Sequential bottomonium suppression in PbPb collisions probes the quark-gluon plasma, while pPb measurements constrain cold nuclear matter effects. Multi-quarkonium production explores multi-parton interactions and exotic-hadron structure. Together, these results establish heavy quarkonium as a laboratory for QCD and a probe for physics beyond the Standard Model, while highlighting unresolved theoretical challenges. Future programs at the LHC, Belle II, and BESIII promise greater precision and new discoveries.

hep-ex↗

ACFD: Asymmetric Cartoon Face Detector

Cartoon face detection is a more challenging task than human face detection due to many difficult scenarios is involved. Aiming at the characteristics of cartoon faces, such as huge differences within the intra-faces, in this paper, we propose an asymmetric cartoon face detector, named ACFD. Specifically, it consists of the following modules: a novel backbone VoVNetV3 comprised of several asymmetric one-shot aggregation modules (AOSA), asymmetric bi-directional feature pyramid network (ABi-FPN), dynamic anchor match strategy (DAM) and the corresponding margin binary classification loss (MBC). In particular, to generate features with diverse receptive fields, multi-scale pyramid features are extracted by VoVNetV3, and then fused and enhanced simultaneously by ABi-FPN for handling the faces in some extreme poses and have disparate aspect ratios. Besides, DAM is used to match enough high-quality anchors for each face, and MBC is for the strong power of discrimination. With the effectiveness of these modules, our ACFD achieves the 1st place on the detection track of 2020 iCartoon Face Challenge under the constraints of model size 200MB, inference time 50ms per image, and without any pretrained models.

cs.CV↗

Fast Learning of Temporal Action Proposal via Dense Boundary Generator

Generating temporal action proposals remains a very challenging problem, where the main issue lies in predicting precise temporal proposal boundaries and reliable action confidence in long and untrimmed real-world videos. In this paper, we propose an efficient and unified framework to generate temporal action proposals named Dense Boundary Generator (DBG), which draws inspiration from boundary-sensitive methods and implements boundary classification and action completeness regression for densely distributed proposals. In particular, the DBG consists of two modules: Temporal boundary classification (TBC) and Action-aware completeness regression (ACR). The TBC aims to provide two temporal boundary confidence maps by low-level two-stream features, while the ACR is designed to generate an action completeness score map by high-level action-aware features. Moreover, we introduce a dual stream BaseNet (DSB) to encode RGB and optical flow information, which helps to capture discriminative boundary and actionness features. Extensive experiments on popular benchmarks ActivityNet-1.3 and THUMOS14 demonstrate the superiority of DBG over the state-of-the-art proposal generator (e.g., MGG and BMN). Our code will be made available upon publication.

cs.CV↗