Searcharxiv⌕ Search

arXiv subjects

Jianfeng Zhang

Publications and source records attributed to Jianfeng Zhang.

At least 91 records · Page 5Linked to original sources
arXiv2023

Continual Learning via Manifold Expansion Replay

In continual learning, the learner learns multiple tasks in sequence, with data being acquired only once for each task. Catastrophic forgetting is a major challenge to continual learning. To reduce forgetting, some existing rehearsal-based methods use episodic memory to replay samples of previous tasks. However, in the process of knowledge integration when learning a new task, this strategy also suffers from catastrophic forgetting due to an imbalance between old and new knowledge. To address this problem, we propose a novel replay strategy called Manifold Expansion Replay (MaER). We argue that expanding the implicit manifold of the knowledge representation in the episodic memory helps to improve the robustness and expressiveness of the model. To this end, we propose a greedy strategy to keep increasing the diameter of the implicit manifold represented by the knowledge in the buffer during memory management. In addition, we introduce Wasserstein distance instead of cross entropy as distillation loss to preserve previous knowledge. With extensive experimental validation on MNIST, CIFAR10, CIFAR100, and TinyImageNet, we show that the proposed method significantly improves the accuracy in continual learning setup, outperforming the state of the arts.

cs.LG↗
arXiv2023

GETAvatar: Generative Textured Meshes for Animatable Human Avatars

We study the problem of 3D-aware full-body human generation, aiming at creating animatable human avatars with high-quality textures and geometries. Generally, two challenges remain in this field: i) existing methods struggle to generate geometries with rich realistic details such as the wrinkles of garments; ii) they typically utilize volumetric radiance fields and neural renderers in the synthesis process, making high-resolution rendering non-trivial. To overcome these problems, we propose GETAvatar, a Generative model that directly generates Explicit Textured 3D meshes for animatable human Avatar, with photo-realistic appearance and fine geometric details. Specifically, we first design an articulated 3D human representation with explicit surface modeling, and enrich the generated humans with realistic surface details by learning from the 2D normal maps of 3D scan data. Second, with the explicit mesh representation, we can use a rasterization-based renderer to perform surface rendering, allowing us to achieve high-resolution image generation efficiently. Extensive experiments demonstrate that GETAvatar achieves state-of-the-art performance on 3D-aware human generation both in appearance and geometry quality. Notably, GETAvatar can generate images at 512x512 resolution with 17FPS and 1024x1024 resolution with 14FPS, improving upon previous methods by 2x. Our code and models will be available.

cs.CV↗
arXiv2023

Minimal solutions of master equations for extended mean field games

In an extended mean field game the vector field governing the flow of the population can be different from that of the individual player at some mean field equilibrium. This new class strictly includes the standard mean field games. It is well known that, without any monotonicity conditions, mean field games typically contain multiple mean field equilibria and the wellposedness of their corresponding master equations fails. In this paper, a partial order for the set of probability measure flows is proposed to compare different mean field equilibria. The minimal and maximal mean field equilibria under this partial order are constructed and satisfy the flow property. The corresponding value functions, however, are in general discontinuous. We thus introduce a notion of weak-viscosity solutions for the master equation and verify that the value functions are indeed weak-viscosity solutions. Moreover, a comparison principle for weak-viscosity semi-solutions is established and thus these two value functions serve as the minimal and maximal weak-viscosity solutions in appropriate sense. In particular, when these two value functions coincide, the value function becomes the unique weak-viscosity solution to the master equation. The novelties of the work persist even when restricted to the standard mean field games.

math.PR↗
arXiv2023

MagicAvatar: Multimodal Avatar Generation and Animation

This report presents MagicAvatar, a framework for multimodal video generation and animation of human avatars. Unlike most existing methods that generate avatar-centric videos directly from multimodal inputs (e.g., text prompts), MagicAvatar explicitly disentangles avatar video generation into two stages: (1) multimodal-to-motion and (2) motion-to-video generation. The first stage translates the multimodal inputs into motion/ control signals (e.g., human pose, depth, DensePose); while the second stage generates avatar-centric video guided by these motion signals. Additionally, MagicAvatar supports avatar animation by simply providing a few images of the target person. This capability enables the animation of the provided human identity according to the specific motion derived from the first stage. We demonstrate the flexibility of MagicAvatar through various applications, including text-guided and video-guided avatar generation, as well as multimodal avatar animation.

cs.GR↗
arXiv2023

MagicEdit: High-Fidelity and Temporally Coherent Video Editing

In this report, we present MagicEdit, a surprisingly simple yet effective solution to the text-guided video editing task. We found that high-fidelity and temporally coherent video-to-video translation can be achieved by explicitly disentangling the learning of content, structure and motion signals during training. This is in contradict to most existing methods which attempt to jointly model both the appearance and temporal representation within a single framework, which we argue, would lead to degradation in per-frame quality. Despite its simplicity, we show that MagicEdit supports various downstream video editing tasks, including video stylization, local editing, video-MagicMix and video outpainting.

cs.CV↗
arXiv2023

A Co-training Approach for Noisy Time Series Learning

In this work, we focus on robust time series representation learning. Our assumption is that real-world time series is noisy and complementary information from different views of the same time series plays an important role while analyzing noisy input. Based on this, we create two views for the input time series through two different encoders. We conduct co-training based contrastive learning iteratively to learn the encoders. Our experiments demonstrate that this co-training approach leads to a significant improvement in performance. Especially, by leveraging the complementary information from different views, our proposed TS-CoT method can mitigate the impact of data noise and corruption. Empirical evaluations on four time series benchmarks in unsupervised and semi-supervised settings reveal that TS-CoT outperforms existing methods. Furthermore, the representations learned by TS-CoT can transfer well to downstream tasks through fine-tuning.

cs.LG↗
arXiv2023

Co-existence of charge density wave and anti-ferromagnetic coupling in the spin-chain compound Ba$_6$Cr$_2$S$_{10}$

Here we have performed detailed first principles calculations for the electronic structure and magnetic properties of Ba$_6$Cr$_2$S$_{10}$ to study the origin of the anti-ferromagnetic exchange interaction between spins on Cr ions for the spin-chain compound Ba$_6$Cr$_2$S$_{10}$ synthesised recently. Most importantly, we have found the co-existence of a charge density wave phase along one line and an anti-ferromagnetic spin chain along another. The dimerization of sulfur atoms loosely bonded with Ba atoms drives the system into an insulating state owing to the formation of charge density wave. Meanwhile, the small size of the effective Hubbard $U$ parameter ($\sim 0.5$ eV) due to electrostatic screening mainly accounts for the anti-ferromagnetic ground state. This co-existence equips us with a platform to tune the charge and spin degrees of freedom independently. Moreover, there exists a next-nearest-neighbouring anti-ferromagnetic interaction along the chain, which could bring forward spin frustration and hence quantum spin liquid.

cond-mat.str-el↗
arXiv2023

Structural, electronic, magnetic properties of Cu-doped lead-apatite Pb$_{10-x}$Cu$_x$(PO$_4$)$_6$O

The recent report of superconductivity in the Cu-doped PbPO compound stimulates the extensive researches on its physical properties. Herein, the detailed atomic and electronic structures of this compound are investigated, which are the necessary information to explain the physical properties, including possible superconductivity. By the first-principles electronic structure calculations, we find that the partial replacement of Pb at $4f$ site by Cu atom, instead of Pb at $6h$ site, plays a crucial role in dominating the electronic state at Fermi energy. The $3d$ electronic orbitals of Cu atom emerge near the Fermi energy and exhibit strong spin-polarization, resulting in the local moment around the doped Cu atom. Particularly, the ground state of Pb$_{10-x}$Cu$_x$(PO$_4$)$_6$O (x = 1) is determined to be a semiconducting phase, in good agreement with the experimental measurements.

cond-mat.supr-con↗
arXiv2023

Is a sophisticated agent always a wise one?

For time inconsistent optimal control problems, a quite popular approach is the equilibrium approach, taken by the sophisticated agents. In this short note we construct a deterministic continuous time example where the unique equilibrium is dominated by another control. Therefore, in this situation it may not be wise to take the equilibrium strategy.

math.OC↗
arXiv2023

Parse and Recall: Towards Accurate Lung Nodule Malignancy Prediction like Radiologists

Lung cancer is a leading cause of death worldwide and early screening is critical for improving survival outcomes. In clinical practice, the contextual structure of nodules and the accumulated experience of radiologists are the two core elements related to the accuracy of identification of benign and malignant nodules. Contextual information provides comprehensive information about nodules such as location, shape, and peripheral vessels, and experienced radiologists can search for clues from previous cases as a reference to enrich the basis of decision-making. In this paper, we propose a radiologist-inspired method to simulate the diagnostic process of radiologists, which is composed of context parsing and prototype recalling modules. The context parsing module first segments the context structure of nodules and then aggregates contextual information for a more comprehensive understanding of the nodule. The prototype recalling module utilizes prototype-based learning to condense previously learned cases as prototypes for comparative analysis, which is updated online in a momentum way during training. Building on the two modules, our method leverages both the intrinsic characteristics of the nodules and the external knowledge accumulated from other nodules to achieve a sound diagnosis. To meet the needs of both low-dose and noncontrast screening, we collect a large-scale dataset of 12,852 and 4,029 nodules from low-dose and noncontrast CTs respectively, each with pathology- or follow-up-confirmed labels. Experiments on several datasets demonstrate that our method achieves advanced screening performance on both low-dose and noncontrast scenarios.

eess.IV↗
arXiv2023

Cubature Method for Stochastic Volterra Integral Equations

In this paper, we introduce the cubature formula for Stochastic Volterra Integral Equations. We first derive the stochastic Taylor expansion in this setting, by utilizing a functional Itô formula, and provide its tail estimates. We then introduce the cubature measure for such equations, and construct it explicitly in some special cases, including a long memory stochastic volatility model. We shall provide the error estimate rigorously. Our numerical examples show that the cubature method is much more efficient than the Euler scheme, provided certain conditions are satisfied.

math.PR↗
arXiv2023

Group Equivariant BEV for 3D Object Detection

Recently, 3D object detection has attracted significant attention and achieved continuous improvement in real road scenarios. The environmental information is collected from a single sensor or multi-sensor fusion to detect interested objects. However, most of the current 3D object detection approaches focus on developing advanced network architectures to improve the detection precision of the object rather than considering the dynamic driving scenes, where data collected from sensors equipped in the vehicle contain various perturbation features. As a result, existing work cannot still tackle the perturbation issue. In order to solve this problem, we propose a group equivariant bird's eye view network (GeqBevNet) based on the group equivariant theory, which introduces the concept of group equivariant into the BEV fusion object detection network. The group equivariant network is embedded into the fused BEV feature map to facilitate the BEV-level rotational equivariant feature extraction, thus leading to lower average orientation error. In order to demonstrate the effectiveness of the GeqBevNet, the network is verified on the nuScenes validation dataset in which mAOE can be decreased to 0.325. Experimental results demonstrate that GeqBevNet can extract more rotational equivariant features in the 3D object detection of the actual road scene and improve the performance of object orientation prediction.

cs.CV↗
arXiv2023

PV3D: A 3D Generative Model for Portrait Video Generation

Recent advances in generative adversarial networks (GANs) have demonstrated the capabilities of generating stunning photo-realistic portrait images. While some prior works have applied such image GANs to unconditional 2D portrait video generation and static 3D portrait synthesis, there are few works successfully extending GANs for generating 3D-aware portrait videos. In this work, we propose PV3D, the first generative framework that can synthesize multi-view consistent portrait videos. Specifically, our method extends the recent static 3D-aware image GAN to the video domain by generalizing the 3D implicit neural representation to model the spatio-temporal space. To introduce motion dynamics to the generation process, we develop a motion generator by stacking multiple motion layers to generate motion features via modulated convolution. To alleviate motion ambiguities caused by camera/human motions, we propose a simple yet effective camera condition strategy for PV3D, enabling both temporal and multi-view consistent video generation. Moreover, PV3D introduces two discriminators for regularizing the spatial and temporal domains to ensure the plausibility of the generated portrait videos. These elaborated designs enable PV3D to generate 3D-aware motion-plausible portrait videos with high-quality appearance and geometry, significantly outperforming prior works. As a result, PV3D is able to support many downstream applications such as animating static portraits and view-consistent video motion editing. Code and models are released at https://showlab.github.io/pv3d.

cs.CV↗
arXiv2023

Ferrimagnetic Regulation of Weyl Fermions in a Noncentrosymmetric Magnetic Weyl Semimetal

The study of interaction between electromagnetism and elementary particles is a long-standing topic in physics. Likewise, the connection between particle physics and emergent phenomena in condensed matter physics is a recurring theme and condensed matter physics has often provided a platform for investigating the interplay between particles and fields in cases that have not been observed in high-energy physics, so far. Here, using angle-resolved photoemission spectroscopy, we provide a new example of this by visualizing the electronic structure of a noncentrosymmetric magnetic Weyl semimetal candidate NdAlSi in both the paramagnetic and ferrimagnetic states. We observe surface Fermi arcs and bulk Weyl fermion dispersion as well as the regulation of Weyl fermions by ferrimagnetism. Our results establish NdAlSi as a magnetic Weyl semimetal and provide the first experimental observation of ferrimagnetic regulation of Weyl fermions in condensed matter.

cond-mat.str-el↗
arXiv2023

OmniAvatar: Geometry-Guided Controllable 3D Head Synthesis

We present OmniAvatar, a novel geometry-guided 3D head synthesis model trained from in-the-wild unstructured images that is capable of synthesizing diverse identity-preserved 3D heads with compelling dynamic details under full disentangled control over camera poses, facial expressions, head shapes, articulated neck and jaw poses. To achieve such high level of disentangled control, we first explicitly define a novel semantic signed distance function (SDF) around a head geometry (FLAME) conditioned on the control parameters. This semantic SDF allows us to build a differentiable volumetric correspondence map from the observation space to a disentangled canonical space from all the control parameters. We then leverage the 3D-aware GAN framework (EG3D) to synthesize detailed shape and appearance of 3D full heads in the canonical space, followed by a volume rendering step guided by the volumetric correspondence map to output into the observation space. To ensure the control accuracy on the synthesized head shapes and expressions, we introduce a geometry prior loss to conform to head SDF and a control loss to conform to the expression code. Further, we enhance the temporal realism with dynamic details conditioned upon varying expressions and joint poses. Our model can synthesize more preferable identity-preserved 3D heads with compelling dynamic details compared to the state-of-the-art methods both qualitatively and quantitatively. We also provide an ablation study to justify many of our system design choices.

cs.CV↗
arXiv2023

AgileGAN3D: Few-Shot 3D Portrait Stylization by Augmented Transfer Learning

While substantial progresses have been made in automated 2D portrait stylization, admirable 3D portrait stylization from a single user photo remains to be an unresolved challenge. One primary obstacle here is the lack of high quality stylized 3D training data. In this paper, we propose a novel framework \emph{AgileGAN3D} that can produce 3D artistically appealing and personalized portraits with detailed geometry. New stylization can be obtained with just a few (around 20) unpaired 2D exemplars. We achieve this by first leveraging existing 2D stylization capabilities, \emph{style prior creation}, to produce a large amount of augmented 2D style exemplars. These augmented exemplars are generated with accurate camera pose labels, as well as paired real face images, which prove to be critical for the downstream 3D stylization task. Capitalizing on the recent advancement of 3D-aware GAN models, we perform \emph{guided transfer learning} on a pretrained 3D GAN generator to produce multi-view-consistent stylized renderings. In order to achieve 3D GAN inversion that can preserve subject's identity well, we incorporate \emph{multi-view consistency loss} in the training of our encoder. Our pipeline demonstrates strong capability in turning user photos into a diverse range of 3D artistic portraits. Both qualitative results and quantitative evaluations have been conducted to show the superior performance of our method. Code and pretrained models will be released for reproduction purpose.

cs.CV↗
arXiv2023

Inducing Neural Collapse in Deep Long-tailed Learning

Although deep neural networks achieve tremendous success on various classification tasks, the generalization ability drops sheer when training datasets exhibit long-tailed distributions. One of the reasons is that the learned representations (i.e. features) from the imbalanced datasets are less effective than those from balanced datasets. Specifically, the learned representation under class-balanced distribution will present the Neural Collapse (NC) phenomena. NC indicates the features from the same category are close to each other and from different categories are maximally distant, showing an optimal linear separable state of classification. However, the pattern differs on imbalanced datasets and is partially responsible for the reduced performance of the model. In this work, we propose two explicit feature regularization terms to learn high-quality representation for class-imbalanced data. With the proposed regularization, NC phenomena will appear under the class-imbalanced distribution, and the generalization ability can be significantly improved. Our method is easily implemented, highly effective, and can be plugged into most existing methods. The extensive experimental results on widely-used benchmarks show the effectiveness of our method

cs.LG↗
arXiv2023

Dynamic programming equation for the mean field optimal stopping problem

We study the optimal stopping problem of McKean-Vlasov diffusions when the criterion is a function of the law of the stopped process. A remarkable new feature in this setting is that the stopping time also impacts the dynamics of the stopped process through the dependence of the coefficients on the law. The mean field stopping problem is introduced in weak formulation in terms of the joint marginal law of the stopped underlying process and the survival process. This specification satisfies a dynamic programming principle. The corresponding dynamic programming equation is an obstacle problem on the Wasserstein space, and is obtained by means of a general Itô formula for flows of marginal laws of càdlàg semimartingales. Our verification result characterizes the nature of optimal stopping policies, highlighting the crucial need to randomized stopping. The effectiveness of our dynamic programming equation is illustrated by various examples including the mean-variance optimal stopping problem.

math.PR↗