no code implementations • ICML 2020 • Kun Xu, Chongxuan Li, Jun Zhu, Bo Zhang
There are existing efforts that model the training dynamics of GANs in the parameter space but the analysis cannot directly motivate practically effective stabilizing methods.
1 code implementation • 30 Apr 2024 • Luxi Chen, Zhengyi Wang, Chongxuan Li, Tingting Gao, Hang Su, Jun Zhu
In this paper, we introduce score-based iterative reconstruction (SIR), an efficient and general algorithm for 3D generation with a multi-view score-based diffusion model.
1 code implementation • 24 Apr 2024 • Kaiwen Xue, Yuhao Zhou, Shen Nie, Xu Min, Xiaolu Zhang, Jun Zhou, Chongxuan Li
Bayesian flow networks (BFNs) iteratively refine the parameters, instead of the samples in diffusion models (DMs), of distributions at various noise levels through Bayesian inference.
no code implementations • 8 Mar 2024 • Zhengyi Wang, Yikai Wang, Yifei Chen, Chendong Xiang, Shuo Chen, Dajiang Yu, Chongxuan Li, Hang Su, Jun Zhu
In this work, we present the Convolutional Reconstruction Model (CRM), a high-fidelity feed-forward single image-to-3D generative model.
1 code implementation • 26 Feb 2024 • Yijing Liu, Chao Du, Tianyu Pang, Chongxuan Li, Wei Chen, Min Lin
Recent research has made significant progress in optimizing diffusion models for specific downstream objectives, which is an important pursuit in fields such as graph generation for drug design.
no code implementations • 2 Nov 2023 • Shen Nie, Hanzhong Allan Guo, Cheng Lu, Yuhao Zhou, Chenyu Zheng, Chongxuan Li
We present a unified probabilistic formulation for diffusion-based image editing, where a latent variable is edited in a task-specific manner and generally deviates from the corresponding marginal distribution induced by the original stochastic or ordinary differential equation (SDE or ODE).
1 code implementation • 17 Oct 2023 • Siqi Kou, Lei Gan, Dequan Wang, Chongxuan Li, Zhijie Deng
In particular, we derive a novel uncertainty iteration principle to characterize the uncertainty dynamics in diffusion, and leverage the last-layer Laplace approximation for efficient Bayesian inference.
2 code implementations • 4 Oct 2023 • Xiangming Gu, Chao Du, Tianyu Pang, Chongxuan Li, Min Lin, Ye Wang
Looking into this, we first observe that memorization behaviors tend to occur on smaller-sized datasets, which motivates our definition of effective model memorization (EMM), a metric measuring the maximum size of training data at which a learned diffusion model approximates its theoretical optimum.
1 code implementation • 15 Aug 2023 • XiMing Xing, Chuang Wang, Haitao Zhou, Zhihao Hu, Chongxuan Li, Dong Xu, Qian Yu
In the full-control inversion process, we propose an appearance-energy function to control the color and texture of the final generated photo. Importantly, our Inversion-by-Inversion pipeline is training-free and can accept different types of exemplars for color and texture control.
no code implementations • 2 Jul 2023 • Yidong Ouyang, Liyan Xie, Chongxuan Li, Guang Cheng
The diffusion model has shown remarkable performance in modeling data distributions and synthesizing data.
1 code implementation • 26 May 2023 • Min Zhao, Rongzhen Wang, Fan Bao, Chongxuan Li, Jun Zhu
This paper presents \emph{ControlVideo} for text-driven video editing -- generating a video that aligns with a given text while preserving the structure of the source video.
1 code implementation • NeurIPS 2023 • Yunqing Zhao, Tianyu Pang, Chao Du, Xiao Yang, Chongxuan Li, Ngai-Man Cheung, Min Lin
Large vision-language models (VLMs) such as GPT-4 have achieved unprecedented performance in response generation, especially with visual inputs, enabling more creative and adaptable interaction than large language models such as ChatGPT.
2 code implementations • NeurIPS 2023 • Zhengyi Wang, Cheng Lu, Yikai Wang, Fan Bao, Chongxuan Li, Hang Su, Jun Zhu
In comparison, VSD works well with various CFG weights as ancestral sampling from diffusion models and simultaneously improves the diversity and sample quality with a common CFG weight (i. e., $7. 5$).
1 code implementation • 9 May 2023 • Guoqiang Wu, Chongxuan Li, Yilong Yin
We theoretically identify a critical factor of the dataset affecting the generalization bounds: \emph{the label-wise class imbalance}.
3 code implementations • 25 Apr 2023 • Cheng Lu, Huayu Chen, Jianfei Chen, Hang Su, Chongxuan Li, Jun Zhu
The main challenge for this setting is that the intermediate guidance during the diffusion sampling procedure, which is jointly defined by the sampling distribution and the energy function, is unknown and is hard to estimate.
1 code implementation • 31 Mar 2023 • Chendong Xiang, Fan Bao, Chongxuan Li, Hang Su, Jun Zhu
Large-scale diffusion models like Stable Diffusion are powerful and find various real-world applications while customizing such models by fine-tuning is both memory and time inefficient.
3 code implementations • 12 Mar 2023 • Fan Bao, Shen Nie, Kaiwen Xue, Chongxuan Li, Shi Pu, Yaole Wang, Gang Yue, Yue Cao, Hang Su, Jun Zhu
Inspired by the unified view, UniDiffuser learns all distributions simultaneously with a minimal modification to the original diffusion model -- perturbs data in all modalities instead of a single modality, inputs individual timesteps in different modalities, and predicts the noise of all modalities instead of a single modality.
2 code implementations • NeurIPS 2023 • Zebin You, Yong Zhong, Fan Bao, Jiacheng Sun, Chongxuan Li, Jun Zhu
In an effort to further advance semi-supervised generative and classification tasks, we propose a simple yet effective training strategy called dual pseudo training (DPT), built upon strong semi-supervised learners and diffusion models.
1 code implementation • 5 Feb 2023 • Chenyu Zheng, Guoqiang Wu, Fan Bao, Yue Cao, Chongxuan Li, Jun Zhu
Theoretically, the paper considers the surrogate loss instead of the zero-one loss in analyses and generalizes the classical results from binary cases to multiclass ones.
no code implementations • 1 Dec 2022 • Fan Bao, Chongxuan Li, Jiacheng Sun, Jun Zhu
Extensive empirical evidence demonstrates that conditional generative models are easier to train and perform better than unconditional ones by exploiting the labels of data.
1 code implementation • 21 Nov 2022 • Luping Liu, Yi Ren, Xize Cheng, Rongjie Huang, Chongxuan Li, Zhou Zhao
In this paper, we introduce a new perceptron bias assumption that suggests discriminator models are more sensitive to certain features of the input, leading to the overconfidence problem.
1 code implementation • 2 Nov 2022 • Cheng Lu, Yuhao Zhou, Fan Bao, Jianfei Chen, Chongxuan Li, Jun Zhu
The commonly-used fast sampler for guided sampling is DDIM, a first-order diffusion ODE solver that generally needs 100 to 250 steps for high-quality samples.
2 code implementations • 30 Sep 2022 • Fan Bao, Min Zhao, Zhongkai Hao, Peiyao Li, Chongxuan Li, Jun Zhu
Inverse molecular design is critical in material science and drug discovery, where the generated molecules should satisfy certain desirable properties.
3 code implementations • CVPR 2023 • Fan Bao, Shen Nie, Kaiwen Xue, Yue Cao, Chongxuan Li, Hang Su, Jun Zhu
We evaluate U-ViT in unconditional and class-conditional image generation, as well as text-to-image generation tasks, where U-ViT is comparable if not superior to a CNN-based U-Net of a similar size.
Ranked #4 on Text-to-Image Generation on MS COCO
1 code implementation • 30 Aug 2022 • Yong Zhong, Hongtao Liu, Xiaodong Liu, Fan Bao, Weiran Shen, Chongxuan Li
Deep generative models (DGMs) are data-eager because learning a complex model on limited data suffers from a large variance and easily overfits.
1 code implementation • 14 Jul 2022 • Min Zhao, Fan Bao, Chongxuan Li, Jun Zhu
Further, we provide an alternative explanation of the EGSDE as a product of experts, where each of the three experts (corresponding to the SDE and two feature extractors) solely contributes to faithfulness or realism.
Ranked #1 on Image-to-Image Translation on AFHQ (Wild to Dog)
1 code implementation • 17 Jun 2022 • Siyu Wang, Jianfei Chen, Chongxuan Li, Jun Zhu, Bo Zhang
In this work, we propose Integer-only Discrete Flows (IODF), an efficient neural compressor with integer-only arithmetic.
1 code implementation • 16 Jun 2022 • Cheng Lu, Kaiwen Zheng, Fan Bao, Jianfei Chen, Chongxuan Li, Jun Zhu
To fill up this gap, we show that the negative likelihood of the ODE can be bounded by controlling the first, second, and third-order score matching errors; and we further present a novel high-order denoising score matching method to enable maximum likelihood training of score-based diffusion ODEs.
1 code implementation • 15 Jun 2022 • Fan Bao, Chongxuan Li, Jiacheng Sun, Jun Zhu, Bo Zhang
Thus, the generation performance on a subset of timesteps is crucial, which is greatly influenced by the covariance design in DPMs.
2 code implementations • 2 Jun 2022 • Cheng Lu, Yuhao Zhou, Fan Bao, Jianfei Chen, Chongxuan Li, Jun Zhu
In this work, we propose an exact formulation of the solution of diffusion ODEs.
1 code implementation • ICLR 2022 • Liyuan Wang, Xingxing Zhang, Kuo Yang, Longhui Yu, Chongxuan Li, Lanqing Hong, Shifeng Zhang, Zhenguo Li, Yi Zhong, Jun Zhu
In this work, we propose memory replay with data compression (MRDC) to reduce the storage cost of old training samples and thus increase their amount that can be stored in the memory buffer.
2 code implementations • ICLR 2022 • Fan Bao, Chongxuan Li, Jun Zhu, Bo Zhang
In this work, we present a surprising result that both the optimal reverse variance and the corresponding optimal KL divergence of a DPM have analytic forms w. r. t.
1 code implementation • NeurIPS 2021 • Fan Bao, Guoqiang Wu, Chongxuan Li, Jun Zhu, Bo Zhang
Our results can explain some mysterious behaviours of the bilevel programming in practice, for instance, overfitting to the validation set.
no code implementations • NeurIPS 2021 • Guoqiang Wu, Chongxuan Li, Kun Xu, Jun Zhu
Our results show that learning algorithms with the consistent univariate loss have an error bound of $O(c)$ ($c$ is the number of labels), while algorithms with the inconsistent pairwise loss depend on $O(\sqrt{c})$ as shown in prior work.
1 code implementation • ICLR 2021 • Tsung Wei Tsai, Chongxuan Li, Jun Zhu
We present Mixture of Contrastive Experts (MiCE), a unified probabilistic clustering framework that simultaneously exploits the discriminative representations learned by contrastive learning and the semantic structures captured by a latent mixture model.
Ranked #9 on Image Clustering on Imagenet-dog-15
1 code implementation • ICLR 2021 • Cheng Lu, Jianfei Chen, Chongxuan Li, Qiuhao Wang, Jun Zhu
Through theoretical analysis, we show that the function space of ImpFlow is strictly richer than that of ResFlows.
no code implementations • 5 Jan 2021 • Qijun Luo, Zhili Liu, Lanqing Hong, Chongxuan Li, Kuo Yang, Liyuan Wang, Fengwei Zhou, Guilin Li, Zhenguo Li, Jun Zhu
Semi-supervised domain adaptation (SSDA), which aims to learn models in a partially labeled target domain with the assistance of the fully labeled source domain, attracts increasing attention in recent years.
no code implementations • CVPR 2021 • Liyuan Wang, Kuo Yang, Chongxuan Li, Lanqing Hong, Zhenguo Li, Jun Zhu
Continual learning usually assumes the incoming data are fully labeled, which might not be applicable in real applications.
1 code implementation • NeurIPS Workshop ICBINB 2020 • Fan Bao, Kun Xu, Chongxuan Li, Lanqing Hong, Jun Zhu, Bo Zhang
The learning and evaluation of energy-based latent variable models (EBLVMs) without any structural assumptions are highly challenging, because the true posteriors and the partition functions in such models are generally intractable.
1 code implementation • NeurIPS 2020 • Fan Bao, Chongxuan Li, Kun Xu, Hang Su, Jun Zhu, Bo Zhang
This paper presents a bi-level score matching (BiSM) method to learn EBLVMs with general structures by reformulating SM as a bi-level optimization problem.
1 code implementation • NeurIPS 2020 • Tianyu Pang, Kun Xu, Chongxuan Li, Yang song, Stefano Ermon, Jun Zhu
Several machine learning applications involve the optimization of higher-order derivatives (e. g., gradients of gradients) during training, which can be expensive in respect to memory and computation even with automatic differentiation.
1 code implementation • 20 Dec 2019 • Chongxuan Li, Kun Xu, Jiashuo Liu, Jun Zhu, Bo Zhang
It is formulated as a three-player minimax game consisting of a generator, a classifier and a discriminator, and therefore is referred to as Triple Generative Adversarial Network (Triple-GAN).
1 code implementation • 29 Sep 2019 • Kun Xu, Chongxuan Li, Jun Zhu, Bo Zhang
There are existing efforts that model the training dynamics of GANs in the parameter space but the analysis cannot directly motivate practically effective stabilizing methods.
Ranked #37 on Image Generation on CIFAR-10 (Inception score metric)
1 code implementation • NeurIPS 2019 • Kun Xu, Chongxuan Li, Jun Zhu, Bo Zhang
Deep generative models (DGMs) have shown promise in image generation.
no code implementations • 23 May 2019 • Tsung Wei Tsai, Chongxuan Li, Jun Zhu
We consider the learning from noisy labels (NL) problem which emerges in many real-world applications.
no code implementations • ICLR 2020 • Chongxuan Li, Chao Du, Kun Xu, Max Welling, Jun Zhu, Bo Zhang
We propose a black-box algorithm called {\it Adversarial Variational Inference and Learning} (AdVIL) to perform inference and learning on a general Markov random field (MRF).
no code implementations • ICLR 2019 • Chao Du, Kun Xu, Chongxuan Li, Jun Zhu, Bo Zhang
Implicit generative models are difficult to train as no explicit density functions are defined.
1 code implementation • NeurIPS 2018 • Chongxuan Li, Max Welling, Jun Zhu, Bo Zhang
We propose Graphical Generative Adversarial Networks (Graphical-GAN) to model structured data.
no code implementations • NeurIPS 2017 • Jianfei Chen, Chongxuan Li, Yizhong Ru, Jun Zhu
In this paper, we propose population matching discrepancy (PMD) for estimating the distribution distance based on samples, as well as an algorithm to learn the parameters of the distributions using PMD as an objective.
1 code implementation • NeurIPS 2017 • Chongxuan Li, Kun Xu, Jun Zhu, Bo Zhang
Generative Adversarial Nets (GANs) have shown promise in image generation and semi-supervised learning (SSL).
1 code implementation • 21 Dec 2016 • Chao Du, Chongxuan Li, Yin Zheng, Jun Zhu, Bo Zhang
Deep neural networks have shown promise in collaborative filtering (CF).
1 code implementation • 22 Nov 2016 • Chongxuan Li, Jun Zhu, Bo Zhang
Deep generative models (DGMs) are effective on learning multilayered representations of complex data and performing inference of input data by exploring the generative ability.
no code implementations • 24 Apr 2016 • Mengchen Liu, Jiaxin Shi, Zhen Li, Chongxuan Li, Jun Zhu, Shixia Liu
Deep convolutional neural networks (CNNs) have achieved breakthrough performance in many pattern recognition tasks such as image classification.
1 code implementation • 24 Feb 2016 • Chongxuan Li, Jun Zhu, Bo Zhang
Memory units have been widely used to enrich the capabilities of deep networks on capturing long-term dependencies in reasoning and prediction tasks, but little investigation exists on deep generative models (DGMs) which are good at inferring high-level invariant representations from unlabeled data.
2 code implementations • NeurIPS 2015 • Chongxuan Li, Jun Zhu, Tianlin Shi, Bo Zhang
Deep generative models (DGMs) are effective on learning multilayered representations of complex data and performing inference of input data by exploring the generative ability.