arXiv subjects

Weibo Mao

Publications and source records attributed to Weibo Mao.

12 recordsLinked to original sources

arXiv2026

Relaxation Process During Complex Time Evolution In Two-Dimensional Integrable and Chaotic CFTs

We investigate the complex time evolution of a vacuum state with the insertion of a local primary operator in two-dimensional conformal field theories (2d CFTs). This complex time evolution can be considered as a composite process constructed from Lorentzian time evolution and a Euclidean evolution induced by a post-selected measurement. Our main finding is that in the spatially-compact system, this complex time evolution drives the state of the subsystems to those of the primary state with the same conformal dimensions of the inserted operator. Contrary to the compact system, the subsystems of the spatially non-compact system evolve to states that depend on the non-unitary process during a certain time regime. In holographic systems with a compact spatial direction, this process induced by a heavy local operator can correspond to the relaxation from a black hole with an inhomogeneous horizon to that with a uniform one, while in the ones with a non-compact spatial direction, it can correspond to the relaxation to that with a horizon depending on the non-unitary process.

hep-th↗

arXiv2025

Entanglement Dynamics by (Non-)Unitary Local Operator Quenches in a 2D Holographic CFT

In this paper, we investigate the time evolution of entanglement entropy and mutual information for the spatially-infinite systems where we act with a primary operator on the vacuum state and then time-evolve it with the sequence of the Euclidean and Lorentzian time evolutions. Two-dimensional holographic conformal field theories describe the systems under consideration in this paper. The Euclidean time evolution is induced by the Rindler Hamiltonian and behaves as the regulator that tames the divergence induced by the local operator, while the Lorentzian one is induced by the uniform Hamiltonian. Under these time evolutions, we investigate the time ordering effect of the Rindler Euclidean and uniform Lorentzian time evolution operators. Consequently, we find the remarkable differences between those time evolutions are induced by whether those are unitary or non-unitary. Especially, we find that the unitary time evolution induces the late-time logarithmic growth of the entanglement entropy, while the non-unitary time evolution induces the late-time constant behavior. Furthermore, we investigate the dual gravity of the systems under consideration. Especially, we investigate the gravity duals of the systems with the insertion of the heavy primary operator and show that it is a black brane with a spacetime-dependent horizon.

hep-th↗

arXiv2025

Hayden--Preskill Model via Local Quenches

We model the Hayden--Preskill (HP) information recovery protocol in 2d CFTs via local joining quenches. Euclidean path integrals with slits prepare the HP subsystems: the message $M$, its reference $N$, the Page-time black hole $B$, the early radiation $E$, and the late radiation $R$; the remaining black hole after emitting $R$ is denoted as $B'$. The single-slit geometry provides an analytically tractable toy model, while the bounded-slit geometry more closely captures the HP setup. In the free Dirac fermion 2d CFT, the mutual information $I(N\!:\!B')$ shows quasi-particle dynamics with partial or full revivals, whereas that in holographic 2d CFTs, which are expected to be maximally chaotic, exhibit sharp transitions: in the bounded-slit case, when the size of the late radiation becomes comparable to that of the reference $N$, $I(N\!:\!B')$ does vanish at late time, otherwise it remains finite. This contrast between free CFTs and holographic CFTs gives a clear characterization of the HP recovery threshold.

hep-th↗

arXiv2025

Time Ordering Effects and Destruction of Quasiparticles in Two-dimensional Holographic CFTs

In this paper, we investigate the entanglement dynamics induced by a composite operator, defined as the local operator evolved with the time evolution operator constructed of the Euclidean and Lorentzian ones. The systems under consideration are described by two-dimensional holographic conformal field theories (2d holographic CFTs), the theories described by gravity. Then, we find that the time ordering of Euclidean and Lorentzian time evolutions affects the behavior of bipartite entanglement and non-local correlation, especially their late-time ones. We also establish the relation between entanglement entropy and energy-momentum densities. This relation states that the late-time behaviors of bipartite entanglement and non-local correlation strongly depend on the energy distribution. We also examine how the decay of the quasiparticle influences the bipartite entanglement and non-local correlation. Furthermore, we investigate the gravity dual of the systems considered in this paper.

hep-th↗

arXiv2024

Ethical-Lens: Curbing Malicious Usages of Open-Source Text-to-Image Models

The burgeoning landscape of text-to-image models, exemplified by innovations such as Midjourney and DALLE 3, has revolutionized content creation across diverse sectors. However, these advancements bring forth critical ethical concerns, particularly with the misuse of open-source models to generate content that violates societal norms. Addressing this, we introduce Ethical-Lens, a framework designed to facilitate the value-aligned usage of text-to-image tools without necessitating internal model revision. Ethical-Lens ensures value alignment in text-to-image models across toxicity and bias dimensions by refining user commands and rectifying model outputs. Systematic evaluation metrics, combining GPT4-V, HEIM, and FairFace scores, assess alignment capability. Our experiments reveal that Ethical-Lens enhances alignment capabilities to levels comparable with or superior to commercial models like DALLE 3, ensuring user-generated content adheres to ethical standards while maintaining image quality. This study indicates the potential of Ethical-Lens to ensure the sustainable development of open-source text-to-image tools and their beneficial integration into society. Our code is available at https://github.com/yuzhu-cai/Ethical-Lens.

cs.CV↗

arXiv2024

Towards Collaborative Autonomous Driving: Simulation Platform and End-to-End System

Vehicle-to-everything-aided autonomous driving (V2X-AD) has a huge potential to provide a safer driving solution. Despite extensive researches in transportation and communication to support V2X-AD, the actual utilization of these infrastructures and communication resources in enhancing driving performances remains largely unexplored. This highlights the necessity of collaborative autonomous driving: a machine learning approach that optimizes the information sharing strategy to improve the driving performance of each vehicle. This effort necessitates two key foundations: a platform capable of generating data to facilitate the training and testing of V2X-AD, and a comprehensive system that integrates full driving-related functionalities with mechanisms for information sharing. From the platform perspective, we present V2Xverse, a comprehensive simulation platform for collaborative autonomous driving. This platform provides a complete pipeline for collaborative driving. From the system perspective, we introduce CoDriving, a novel end-to-end collaborative driving system that properly integrates V2X communication over the entire autonomous pipeline, promoting driving with shared perceptual information. The core idea is a novel driving-oriented communication strategy. Leveraging this strategy, CoDriving improves driving performance while optimizing communication efficiency. We make comprehensive benchmarks with V2Xverse, analyzing both modular performance and closed-loop driving performance. Experimental results show that CoDriving: i) significantly improves the driving score by 62.49% and drastically reduces the pedestrian collision rate by 53.50% compared to the SOTA end-to-end driving method, and ii) achieves sustaining driving performance superiority over dynamic constraint communication conditions.

cs.CV↗

arXiv2024

Local operator quench induced by two-dimensional inhomogeneous and homogeneous CFT Hamiltonians

We explore non-equilibrium processes in two-dimensional conformal field theories (2d CFTs) due to the growth of operators induced by inhomogeneous and homogeneous Hamiltonians by investigating the time dependence of the partition function, energy density, and entanglement entropy. The non-equilibrium processes considered in this paper are constructed out of the Lorentzian and Euclidean time evolution governed by different Hamiltonians. We explore the effect of the time ordering on entanglement dynamics so that we find that in a free boson CFT and RCFTs, this time ordering does not affect the entanglement entropy, while in the holographic CFTs, it does. Our main finding is that in the holographic CFTs, the non-unitary time evolution induced by the inhomogeneous Hamiltonian can retain the initial state information longer than in the unitary time evolution.

hep-th↗

arXiv2023

Joint-Relation Transformer for Multi-Person Motion Prediction

Multi-person motion prediction is a challenging problem due to the dependency of motion on both individual past movements and interactions with other people. Transformer-based methods have shown promising results on this task, but they miss the explicit relation representation between joints, such as skeleton structure and pairwise distance, which is crucial for accurate interaction modeling. In this paper, we propose the Joint-Relation Transformer, which utilizes relation information to enhance interaction modeling and improve future motion prediction. Our relation information contains the relative distance and the intra-/inter-person physical constraints. To fuse relation and joint information, we design a novel joint-relation fusion layer with relation-aware attention to update both features. Additionally, we supervise the relation information by forecasting future distance. Experiments show that our method achieves a 13.4% improvement of 900ms VIM on 3DPW-SoMoF/RC and 17.8%/12.0% improvement of 3s MPJPE on CMU-Mpcap/MuPoTS-3D dataset.

cs.CV↗

arXiv2023

FusionFormer: A Multi-sensory Fusion in Bird's-Eye-View and Temporal Consistent Transformer for 3D Object Detection

Multi-sensor modal fusion has demonstrated strong advantages in 3D object detection tasks. However, existing methods that fuse multi-modal features require transforming features into the bird's eye view space and may lose certain information on Z-axis, thus leading to inferior performance. To this end, we propose a novel end-to-end multi-modal fusion transformer-based framework, dubbed FusionFormer, that incorporates deformable attention and residual structures within the fusion encoding module. Specifically, by developing a uniform sampling strategy, our method can easily sample from 2D image and 3D voxel features spontaneously, thus exploiting flexible adaptability and avoiding explicit transformation to the bird's eye view space during the feature concatenation process. We further implement a residual structure in our feature encoder to ensure the model's robustness in case of missing an input modality. Through extensive experiments on a popular autonomous driving benchmark dataset, nuScenes, our method achieves state-of-the-art single model performance of 72.6% mAP and 75.1% NDS in the 3D object detection task without test time augmentation.

cs.CV↗

arXiv2023

FusionAD: Multi-modality Fusion for Prediction and Planning Tasks of Autonomous Driving

Building a multi-modality multi-task neural network toward accurate and robust performance is a de-facto standard in perception task of autonomous driving. However, leveraging such data from multiple sensors to jointly optimize the prediction and planning tasks remains largely unexplored. In this paper, we present FusionAD, to the best of our knowledge, the first unified framework that fuse the information from two most critical sensors, camera and LiDAR, goes beyond perception task. Concretely, we first build a transformer based multi-modality fusion network to effectively produce fusion based features. In constrast to camera-based end-to-end method UniAD, we then establish a fusion aided modality-aware prediction and status-aware planning modules, dubbed FMSPnP that take advantages of multi-modality features. We conduct extensive experiments on commonly used benchmark nuScenes dataset, our FusionAD achieves state-of-the-art performance and surpassing baselines on average 15% on perception tasks like detection and tracking, 10% on occupancy prediction accuracy, reducing prediction error from 0.708 to 0.389 in ADE score and reduces the collision rate from 0.31% to only 0.12%.

cs.CV↗

arXiv2023

Leapfrog Diffusion Model for Stochastic Trajectory Prediction

To model the indeterminacy of human behaviors, stochastic trajectory prediction requires a sophisticated multi-modal distribution of future trajectories. Emerging diffusion models have revealed their tremendous representation capacities in numerous generation tasks, showing potential for stochastic trajectory prediction. However, expensive time consumption prevents diffusion models from real-time prediction, since a large number of denoising steps are required to assure sufficient representation ability. To resolve the dilemma, we present LEapfrog Diffusion model (LED), a novel diffusion-based trajectory prediction model, which provides real-time, precise, and diverse predictions. The core of the proposed LED is to leverage a trainable leapfrog initializer to directly learn an expressive multi-modal distribution of future trajectories, which skips a large number of denoising steps, significantly accelerating inference speed. Moreover, the leapfrog initializer is trained to appropriately allocate correlated samples to provide a diversity of predicted future trajectories, significantly improving prediction performances. Extensive experiments on four real-world datasets, including NBA/NFL/SDD/ETH-UCY, show that LED consistently improves performance and achieves 23.7%/21.9% ADE/FDE improvement on NFL. The proposed LED also speeds up the inference 19.3/30.8/24.3/25.1 times compared to the standard diffusion model on NBA/NFL/SDD/ETH-UCY, satisfying real-time inference needs. Code is available at https://github.com/MediaBrain-SJTU/LED.

cs.CV↗

arXiv2022

Remember Intentions: Retrospective-Memory-based Trajectory Prediction

To realize trajectory prediction, most previous methods adopt the parameter-based approach, which encodes all the seen past-future instance pairs into model parameters. However, in this way, the model parameters come from all seen instances, which means a huge amount of irrelevant seen instances might also involve in predicting the current situation, disturbing the performance. To provide a more explicit link between the current situation and the seen instances, we imitate the mechanism of retrospective memory in neuropsychology and propose MemoNet, an instance-based approach that predicts the movement intentions of agents by looking for similar scenarios in the training data. In MemoNet, we design a pair of memory banks to explicitly store representative instances in the training set, acting as prefrontal cortex in the neural system, and a trainable memory addresser to adaptively search a current situation with similar instances in the memory bank, acting like basal ganglia. During prediction, MemoNet recalls previous memory by using the memory addresser to index related instances in the memory bank. We further propose a two-step trajectory prediction system, where the first step is to leverage MemoNet to predict the destination and the second step is to fulfill the whole trajectory according to the predicted destinations. Experiments show that the proposed MemoNet improves the FDE by 20.3%/10.2%/28.3% from the previous best method on SDD/ETH-UCY/NBA datasets. Experiments also show that our MemoNet has the ability to trace back to specific instances during prediction, promoting more interpretability.

cs.CV↗