arXiv subjects

Yizhe Zhu

Publications and source records attributed to Yizhe Zhu.

70 records · Page 4Linked to original sources

arXiv2021

Towards Faster and Stabilized GAN Training for High-fidelity Few-shot Image Synthesis

Training Generative Adversarial Networks (GAN) on high-fidelity images usually requires large-scale GPU-clusters and a vast number of training images. In this paper, we study the few-shot image synthesis task for GAN with minimum computing cost. We propose a light-weight GAN structure that gains superior quality on 1024*1024 resolution. Notably, the model converges from scratch with just a few hours of training on a single RTX-2080 GPU, and has a consistent performance, even with less than 100 training samples. Two technique designs constitute our work, a skip-layer channel-wise excitation module and a self-supervised discriminator trained as a feature-encoder. With thirteen datasets covering a wide variety of image domains (The datasets and code are available at: https://github.com/odegeasslbc/FastGAN-pytorch), we show our model's superior performance compared to the state-of-the-art StyleGAN2, when data and computing budget are limited.

cs.CV↗

arXiv2020

Asymptotic Behavior of a Sequence of Conditional Probability Distributions and the Canonical Ensemble

The probability distribution of a function of a subsystem conditioned on the value of the function of the whole, in the limit when the ratio of their values goes to zero, has a limit law: It equals the unconditioned marginal probability distribution weighted by an exponential factor whose exponent is uniquely determined by the condition. We apply this theorem to explain the canonical equilibrium ensemble of a system in contact with a heat reservoir. Since the theorem only requires analysis at the level of the function of the subsystem and reservoir, it is applicable even without the knowledge of the composition of the reservoir itself, which extends the applicability of the canonical ensemble. Furthermore, we generalize our theorem to a model with strong interaction that contributes an additional term to the exponent, which is beyond the typical case of approximately additive functions. This result is new in both physics and mathematics, as a theory for the Gibbs conditioning principle for strongly correlated systems. A corollary provides a precise formulation of what a temperature bath is in probabilistic term

math-ph↗

arXiv2020

TIME: Text and Image Mutual-Translation Adversarial Networks

Focusing on text-to-image (T2I) generation, we propose Text and Image Mutual-Translation Adversarial Networks (TIME), a lightweight but effective model that jointly learns a T2I generator G and an image captioning discriminator D under the Generative Adversarial Network framework. While previous methods tackle the T2I problem as a uni-directional task and use pre-trained language models to enforce the image--text consistency, TIME requires neither extra modules nor pre-training. We show that the performance of G can be boosted substantially by training it jointly with D as a language model. Specifically, we adopt Transformers to model the cross-modal connections between the image features and word embeddings, and design an annealing conditional hinge loss that dynamically balances the adversarial learning. In our experiments, TIME achieves state-of-the-art (SOTA) performance on the CUB and MS-COCO dataset (Inception Score of 4.91 and Fréchet Inception Distance of 14.3 on CUB), and shows promising performance on MS-COCO on image captioning and downstream vision-language tasks.

cs.CV↗

arXiv2020

Self-Supervised Sketch-to-Image Synthesis

Imagining a colored realistic image from an arbitrarily drawn sketch is one of the human capabilities that we eager machines to mimic. Unlike previous methods that either requires the sketch-image pairs or utilize low-quantity detected edges as sketches, we study the exemplar-based sketch-to-image (s2i) synthesis task in a self-supervised learning manner, eliminating the necessity of the paired sketch data. To this end, we first propose an unsupervised method to efficiently synthesize line-sketches for general RGB-only datasets. With the synthetic paired-data, we then present a self-supervised Auto-Encoder (AE) to decouple the content/style features from sketches and RGB-images, and synthesize images that are both content-faithful to the sketches and style-consistent to the RGB-images. While prior works employ either the cycle-consistence loss or dedicated attentional modules to enforce the content/style fidelity, we show AE's superior performance with pure self-supervisions. To further improve the synthesis quality in high resolution, we also leverage an adversarial network to refine the details of synthetic images. Extensive experiments on 1024*1024 resolution demonstrate a new state-of-art-art performance of the proposed model on CelebA-HQ and Wiki-Art datasets. Moreover, with the proposed sketch generator, the model shows a promising performance on style mixing and style transfer, which require synthesized images to be both style-consistent and semantically meaningful. Our code is available on https://github.com/odegeasslbc/Self-Supervised-Sketch-to-Image-Synthesis-PyTorch, and please visit https://create.playform.io/my-projects?mode=sketch for an online demo of our model.

cs.CV↗

arXiv2020

Eigenvalues of the non-backtracking operator detached from the bulk

We describe the non-backtracking spectrum of a stochastic block model with connection probabilities $p_{\mathrm{in}}, p_{\mathrm{out}} = ω(\log n)/n$. In this regime we answer a question posed in Dall'Amico and al. (2019) regarding the existence of a real eigenvalue `inside' the bulk, close to the location $\frac{p_{\mathrm{in}}+ p_{\mathrm{out}}}{p_{\mathrm{in}}- p_{\mathrm{out}}}$. We also introduce a variant of the Bauer-Fike theorem well suited for perturbations of quadratic eigenvalue problems, and which could be of independent interest.

math.PR↗

arXiv2020

S3VAE: Self-Supervised Sequential VAE for Representation Disentanglement and Data Generation

We propose a sequential variational autoencoder to learn disentangled representations of sequential data (e.g., videos and audios) under self-supervision. Specifically, we exploit the benefits of some readily accessible supervisory signals from input data itself or some off-the-shelf functional models and accordingly design auxiliary tasks for our model to utilize these signals. With the supervision of the signals, our model can easily disentangle the representation of an input sequence into static factors and dynamic factors (i.e., time-invariant and time-varying parts). Comprehensive experiments across videos and audios verify the effectiveness of our model on representation disentanglement and generation of sequential data, and demonstrate that, our model with self-supervision performs comparable to, if not better than, the fully-supervised model with ground truth labels, and outperforms state-of-the-art unsupervised models by a large margin.

cs.CV↗

arXiv2020

OOGAN: Disentangling GAN with One-Hot Sampling and Orthogonal Regularization

Exploring the potential of GANs for unsupervised disentanglement learning, this paper proposes a novel GAN-based disentanglement framework with One-Hot Sampling and Orthogonal Regularization (OOGAN). While previous works mostly attempt to tackle disentanglement learning through VAE and seek to implicitly minimize the Total Correlation (TC) objective with various sorts of approximation methods, we show that GANs have a natural advantage in disentangling with an alternating latent variable (noise) sampling method that is straightforward and robust. Furthermore, we provide a brand-new perspective on designing the structure of the generator and discriminator, demonstrating that a minor structural change and an orthogonal regularization on model weights entails an improved disentanglement. Instead of experimenting on simple toy datasets, we conduct experiments on higher-resolution images and show that OOGAN greatly pushes the boundary of unsupervised disentanglement.

cs.CV↗

arXiv2020

Exact Recovery in the Hypergraph Stochastic Block Model: a Spectral Algorithm

We consider the exact recovery problem in the hypergraph stochastic block model (HSBM) with $k$ blocks of equal size. More precisely, we consider a random $d$-uniform hypergraph $H$ with $n$ vertices partitioned into $k$ clusters of size $s = n / k$. Hyperedges $e$ are added independently with probability $p$ if $e$ is contained within a single cluster and $q$ otherwise, where $0 \leq q < p \leq 1$. We present a spectral algorithm which recovers the clusters exactly with high probability, given mild conditions on $n, k, p, q$, and $d$. Our algorithm is based on the adjacency matrix of $H$, which is a symmetric $n \times n$ matrix whose $(u, v)$-th entry is the number of hyperedges containing both $u$ and $v$. To the best of our knowledge, our algorithm is the first to guarantee exact recovery when the number of clusters $k=Θ(\sqrt{n})$.

cs.LG↗

arXiv2019

Semantic-Guided Multi-Attention Localization for Zero-Shot Learning

Zero-shot learning extends the conventional object classification to the unseen class recognition by introducing semantic representations of classes. Existing approaches predominantly focus on learning the proper mapping function for visual-semantic embedding, while neglecting the effect of learning discriminative visual features. In this paper, we study the significance of the discriminative region localization. We propose a semantic-guided multi-attention localization model, which automatically discovers the most discriminative parts of objects for zero-shot learning without any human annotations. Our model jointly learns cooperative global and local features from the whole object as well as the detected parts to categorize objects based on semantic descriptions. Moreover, with the joint supervision of embedding softmax loss and class-center triplet loss, the model is encouraged to learn features with high inter-class dispersion and intra-class compactness. Through comprehensive experiments on three widely used zero-shot learning benchmarks, we show the efficacy of the multi-attention localization and our proposed approach improves the state-of-the-art results by a considerable margin.

cs.CV↗

arXiv2019

Learning Feature-to-Feature Translator by Alternating Back-Propagation for Generative Zero-Shot Learning

We investigate learning feature-to-feature translator networks by alternating back-propagation as a general-purpose solution to zero-shot learning (ZSL) problems. It is a generative model-based ZSL framework. In contrast to models based on generative adversarial networks (GAN) or variational autoencoders (VAE) that require auxiliary networks to assist the training, our model consists of a single conditional generator that maps class-level semantic features and Gaussian white noise vector accounting for instance-level latent factors to visual features, and is trained by maximum likelihood estimation. The training process is a simple yet effective alternating back-propagation process that iterates the following two steps: (i) the inferential back-propagation to infer the latent factors of each observed example, and (ii) the learning back-propagation to update the model parameters. We show that, with slight modifications, our model is capable of learning from incomplete visual features for ZSL. We conduct extensive comparisons with existing generative ZSL methods on five benchmarks, demonstrating the superiority of our method in not only ZSL performance but also convergence speed and computational cost. Specifically, our model outperforms the existing state-of-the-art methods by a remarkable margin up to 3.1% and 4.0% in ZSL and generalized ZSL settings, respectively.

cs.CV↗

arXiv2019

A Graphon Approach to Limiting Spectral Distributions of Wigner-type Matrices

We present a new approach, based on graphon theory, to finding the limiting spectral distributions of general Wigner-type matrices. This approach determines the moments of the limiting measures and the equations of their Stieltjes transforms explicitly with weaker assumptions on the convergence of variance profiles than previous results. As applications, we give a new proof of the semicircle law for generalized Wigner matrices and determine the limiting spectral distributions for three sparse inhomogeneous random graph models with sparsity $ω(1/n)$: inhomogeneous random graphs with roughly equal expected degrees, $W$-random graphs and stochastic block models with a growing number of blocks. Furthermore, we show our theorems can be applied to random Gram matrices with a variance profile for which we can find the limiting spectral distributions under weaker assumptions than previous results.

math.PR↗

arXiv2019

Sparse General Wigner-type Matrices: Local Law and Eigenvector Delocalization

We prove a local law and eigenvector delocalization for general Wigner-type matrices. Our methods allow us to get the best possible interval length and optimal eigenvector delocalization in the dense case, and the first results of such kind for the sparse case down to $p=\frac{g(n)\log n}{n}$ with $g(n)\to\infty$. We specialize our results to the case of the Stochastic Block Model, and we also obtain a local law for the case when the number of classes is unbounded.

math.PR↗

arXiv2018

CU-Net: Coupled U-Nets

We design a new connectivity pattern for the U-Net architecture. Given several stacked U-Nets, we couple each U-Net pair through the connections of their semantic blocks, resulting in the coupled U-Nets (CU-Net). The coupling connections could make the information flow more efficiently across U-Nets. The feature reuse across U-Nets makes each U-Net very parameter efficient. We evaluate the coupled U-Nets on two benchmark datasets of human pose estimation. Both the accuracy and model parameter number are compared. The CU-Net obtains comparable accuracy as state-of-the-art methods. However, it only has at least 60% fewer parameters than other approaches.

cs.CV↗

arXiv2018

A Generative Adversarial Approach for Zero-Shot Learning from Noisy Texts

Most existing zero-shot learning methods consider the problem as a visual semantic embedding one. Given the demonstrated capability of Generative Adversarial Networks(GANs) to generate images, we instead leverage GANs to imagine unseen categories from text descriptions and hence recognize novel classes with no examples being seen. Specifically, we propose a simple yet effective generative model that takes as input noisy text descriptions about an unseen class (e.g.Wikipedia articles) and generates synthesized visual features for this class. With added pseudo data, zero-shot learning is naturally converted to a traditional classification problem. Additionally, to preserve the inter-class discrimination of the generated features, a visual pivot regularization is proposed as an explicit supervision. Unlike previous methods using complex engineered regularizers, our approach can suppress the noise well without additional regularization. Empirically, we show that our method consistently outperforms the state of the art on the largest available benchmarks on Text-based Zero-shot Learning.

cs.CV↗

arXiv2017

A Multilayer-Based Framework for Online Background Subtraction with Freely Moving Cameras

The exponentially increasing use of moving platforms for video capture introduces the urgent need to develop the general background subtraction algorithms with the capability to deal with the moving background. In this paper, we propose a multilayer-based framework for online background subtraction for videos captured by moving cameras. Unlike the previous treatments of the problem, the proposed method is not restricted to binary segmentation of background and foreground, but formulates it as a multi-label segmentation problem by modeling multiple foreground objects in different layers when they appear simultaneously in the scene. We assign an independent processing layer to each foreground object, as well as the background, where both motion and appearance models are estimated, and a probability map is inferred using a Bayesian filtering framework. Finally, Multi-label Graph-cut on Markov Random Field is employed to perform pixel-wise labeling. Extensive evaluation results show that the proposed method outperforms state-of-the-art methods on challenging video sequences.

cs.CV↗

arXiv2017

Link the head to the "beak": Zero Shot Learning from Noisy Text Description at Part Precision

In this paper, we study learning visual classifiers from unstructured text descriptions at part precision with no training images. We propose a learning framework that is able to connect text terms to its relevant parts and suppress connections to non-visual text terms without any part-text annotations. For instance, this learning process enables terms like "beak" to be sparsely linked to the visual representation of parts like head, while reduces the effect of non-visual terms like "migrate" on classifier prediction. Images are encoded by a part-based CNN that detect bird parts and learn part-specific representation. Part-based visual classifiers are predicted from text descriptions of unseen visual classifiers to facilitate classification without training images (also known as zero-shot recognition). We performed our experiments on CUBirds 2011 dataset and improves the state-of-the-art text-based zero-shot recognition results from 34.7\% to 43.6\%. We also created large scale benchmarks on North American Bird Images augmented with text descriptions, where we also show that our approach outperforms existing methods. Our code, data, and models are publically available.

cs.CV↗