Bingchen Liu

Guangyuan Dong, Ziwei Hong, Shenghao Liu et al.

Although Multimodal Sentiment Analysis (MSA) effectively leverages rich information from language, visual, and acoustic modalities, existing methods still face two core challenges: 1) static conflict suppression mechanisms fail to adapt to dynamic variations across samples, and 2) the inherent sentimental bias within the language modality, which can misguide learning from other modalities, remains entangled. To this end, we propose a Dynamic Multimodal Causal Disentanglement and Adaptive Fusion Framework (MCAF). Its cornerstone is the Multi-Granularity Causal Dynamic Router and a Conditional Diffusion Denoising Module. First, we introduce a causal intervention module based on the information bottleneck principle, which builds a Structural Causal Model to disentangle sentimental bias from language features, yielding a "de-confounded" language representation as a pure guiding signal. Second, we devise a Dynamic Multimodal Router that evaluates the interaction states (complementary, conflicting, or redundant) among visual, acoustic, and de-confounded language signals in real-time across three levels: feature, temporal, and modality, then adaptively allocates weights and routes information flow for fine-grained regulation. Finally, a lightweight Conditional Diffusion Denoising Module performs iterative denoising on the fused joint representation to explicitly filter out residual irrelevant information, generating a robust hyper-modality representation. Extensive experiments on the CMU-MOSI and CMU-MOSEI benchmarks show that MCAF sets new state-of-the-art on key classification metrics, achieving an Acc-2/F1 of 86.52%/86.51% on MOSI and 86.72%/86.65% on MOSEI, while remaining highly competitive on others. Comprehensive analyses and visualizations further validate its efficacy in dynamically perceiving interactions, disentangling bias, and enhancing interpretability.

Kunpeng Song, Ligong Han, Bingchen Liu et al.

Can a text-to-image diffusion model be used as a training objective for adapting a GAN generator to another domain? In this paper, we show that the classifier-free guidance can be leveraged as a critic and enable generators to distill knowledge from large-scale text-to-image diffusion models. Generators can be efficiently shifted into new domains indicated by text prompts without access to groundtruth samples from target domains. We demonstrate the effectiveness and controllability of our method through extensive experiments. Although not trained to minimize CLIP loss, our model achieves equally high CLIP scores and significantly lower FID than prior work on short prompts, and outperforms the baseline qualitatively and quantitatively on long and complicated prompts. To our best knowledge, the proposed method is the first attempt at incorporating large-scale pre-trained diffusion models and distillation sampling for text-driven image generator domain adaptation and gives a quality previously beyond possible. Moreover, we extend our work to 3D-aware style-based generators and DreamBooth guidance.

Yufan Zhou, Bingchen Liu, Yizhe Zhu et al.

We present Corgi, a novel method for text-to-image generation. Corgi is based on our proposed shifted diffusion model, which achieves better image embedding generation from input text. Unlike the baseline diffusion model used in DALL-E 2, our method seamlessly encodes prior knowledge of the pre-trained CLIP model in its diffusion process by designing a new initialization distribution and a new transition step of the diffusion. Compared to the strong DALL-E 2 baseline, our method performs better in generating image embedding from the text in terms of both efficiency and effectiveness, resulting in better text-to-image generation. Extensive large-scale experiments are conducted and evaluated in terms of both quantitative measures and human evaluation, indicating a stronger generation ability of our method compared to existing ones. Furthermore, our model enables semi-supervised and language-free training for text-to-image generation, where only part or none of the images in the training dataset have an associated caption. Trained with only 1.7% of the images being captioned, our semi-supervised model obtains FID results comparable to DALL-E 2 on zero-shot text-to-image generation evaluated on MS-COCO. Corgi also achieves new state-of-the-art results across different datasets on downstream language-free text-to-image generation tasks, outperforming the previous method, Lafite, by a large margin.

Bingchen Liu, Ehsan Akhgari, Alexander Visheratin et al.

We introduce Playground v3 (PGv3), our latest text-to-image model that achieves state-of-the-art (SoTA) performance across multiple testing benchmarks, excels in graphic design abilities and introduces new capabilities. Unlike traditional text-to-image generative models that rely on pre-trained language models like T5 or CLIP text encoders, our approach fully integrates Large Language Models (LLMs) with a novel structure that leverages text conditions exclusively from a decoder-only LLM. Additionally, to enhance image captioning quality-we developed an in-house captioner, capable of generating captions with varying levels of detail, enriching the diversity of text structures. We also introduce a new benchmark CapsBench to evaluate detailed image captioning performance. Experimental results demonstrate that PGv3 excels in text prompt adherence, complex reasoning, and accurate text rendering. User preference studies indicate the super-human graphic design ability of our model for common design applications, such as stickers, posters, and logo designs. Furthermore, PGv3 introduces new capabilities, including precise RGB color control and robust multilingual understanding.

Enze Xie, Junsong Chen, Yuyang Zhao et al.

This paper presents SANA-1.5, a linear Diffusion Transformer for efficient scaling in text-to-image generation. Building upon SANA-1.0, we introduce three key innovations: (1) Efficient Training Scaling: A depth-growth paradigm that enables scaling from 1.6B to 4.8B parameters with significantly reduced computational resources, combined with a memory-efficient 8-bit optimizer. (2) Model Depth Pruning: A block importance analysis technique for efficient model compression to arbitrary sizes with minimal quality loss. (3) Inference-time Scaling: A repeated sampling strategy that trades computation for model capacity, enabling smaller models to match larger model quality at inference time. Through these strategies, SANA-1.5 achieves a text-image alignment score of 0.81 on GenEval, which can be further improved to 0.96 through inference scaling with VILA-Judge, establishing a new SoTA on GenEval benchmark. These innovations enable efficient model scaling across different compute budgets while maintaining high quality, making high-quality image generation more accessible. Our code and pre-trained models are released.

Kunpeng Song, Yizhe Zhu, Bingchen Liu et al.

In this paper, we present MoMA: an open-vocabulary, training-free personalized image model that boasts flexible zero-shot capabilities. As foundational text-to-image models rapidly evolve, the demand for robust image-to-image translation grows. Addressing this need, MoMA specializes in subject-driven personalized image generation. Utilizing an open-source, Multimodal Large Language Model (MLLM), we train MoMA to serve a dual role as both a feature extractor and a generator. This approach effectively synergizes reference image and text prompt information to produce valuable image features, facilitating an image diffusion model. To better leverage the generated features, we further introduce a novel self-attention shortcut method that efficiently transfers image features to an image diffusion model, improving the resemblance of the target object in generated images. Remarkably, as a tuning-free plug-and-play module, our model requires only a single reference image and outperforms existing methods in generating images with high detail fidelity, enhanced identity-preservation and prompt faithfulness. Our work is open-source, thereby providing universal access to these advancements.

Bingchen Liu, Shihao Hou, Weixin Zeng et al.

The construction of large open knowledge bases (OKBs) is integral to many applications in the field of mobile computing. Noun phrases and relational phrases in OKBs often suffer from redundancy and ambiguity, which calls for the investigation on OKB canonicalization. However, in order to meet the requirements of some privacy protection regulations and to ensure the timeliness of the data, the canonicalized OKB often needs to remove some sensitive information or outdated data. The machine unlearning in OKB canonicalization is an excellent solution to the above problem. Current solutions address OKB canonicalization by devising advanced clustering algorithms and using knowledge graph embedding (KGE) to further facilitate the canonicalization process. Effective schemes are urgently needed to fully synergise machine unlearning with clustering and KGE learning. To this end, we put forward a multi-task unlearning framework, namely MulCanon, to tackle machine unlearning problem in OKB canonicalization. Specifically, the noise characteristics in the diffusion model are utilized to achieve the effect of machine unlearning for data in OKB. MulCanon unifies the learning objectives of diffusion model, KGE and clustering algorithms, and adopts a two-step multi-task learning paradigm for training. A thorough experimental study on popular OKB canonicalization datasets validates that MulCanon achieves advanced machine unlearning effects.

Bingchen Liu, Yizhe Zhu, Kunpeng Song et al.

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.

Bingchen Liu, Yizhe Zhu, Kunpeng Song et al.

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.

Bingchen Liu, Yuanyuan Fang

Federated learning (FL) promotes the development and application of artificial intelligence technologies by enabling model sharing and collaboration while safeguarding data privacy. Knowledge graph (KG) embedding representation provides a foundation for knowledge reasoning and applications by mapping entities and relations into vector space. Federated KG embedding enables the utilization of knowledge from diverse client sources while safeguarding the privacy of local data. However, due to demands such as privacy protection and the need to adapt to dynamic data changes, investigations into machine unlearning (MU) have been sparked. However, it is challenging to maintain the performance of KG embedding models while forgetting the influence of specific forgotten data on the model. In this paper, we propose FedDM, a novel framework tailored for machine unlearning in federated knowledge graphs. Leveraging diffusion models, we generate noisy data to sensibly mitigate the influence of specific knowledge on FL models while preserving the overall performance concerning the remaining data. We conduct experimental evaluations on benchmark datasets to assess the efficacy of the proposed model. Extensive experiments demonstrate that FedDM yields promising results in knowledge forgetting.

Naixing Xu, Qian Li, Xu Wang et al.

Knowledge graph (KG) embedding methods map entities and relations into continuous vector spaces, improving performance in tasks like link prediction and question answering. With rising privacy concerns, machine unlearning (MU) has emerged as a critical AI technology, enabling models to eliminate the influence of specific data. Existing MU approaches often rely on data obfuscation and adjustments to training loss but lack generalization across unlearning tasks. This paper introduces MetaEU, a Meta-Learning-Based Knowledge Graph Embedding Unlearning framework. MetaEU leverages meta-learning to unlearn specific embeddings, mitigating their impact while preserving model performance on remaining data. Experiments on benchmark datasets demonstrate its effectiveness in KG embedding unlearning.

Bingchen Liu, Jingchen Li, Yuanyuan Fang et al.

Zero-shot learning (ZL) is crucial for tasks involving unseen categories, such as natural language processing, image classification, and cross-lingual transfer.Current applications often fail to accurately infer and handle new relations orentities involving unseen categories, severely limiting their scalability and prac-ticality in open-domain scenarios. ZL learning faces the challenge of effectivelytransferring semantic information of unseen categories in multi-modal knowledgegraph (MMKG) embedding representation learning. In this paper, we proposeZSLLM, a framework for zero-shot embedding learning of MMKGs using largelanguage models (LLMs). We leverage textual modality information of unseencategories as prompts to fully utilize the reasoning capabilities of LLMs, enablingsemantic information transfer across different modalities for unseen categories.Through model-based learning, the embedding representation of unseen cate-gories in MMKG is enhanced. Extensive experiments conducted on multiplereal-world datasets demonstrate the superiority of our approach compared tostate-of-the-art methods.

Bingchen Liu, Yuanyuan Fang, Naixing Xu et al.

Large language models (LLMs) have garnered significant attention for their superior performance in many knowledge-driven applications on the world wide web.These models are designed to train hundreds of millions or more parameters on large amounts of text data, enabling them to understand and generate naturallanguage effectively. As the superior performance of LLMs becomes apparent,they are increasingly being applied to knowledge graph embedding (KGE) related tasks to improve the processing results. Traditional KGE representation learning methods map entities and relations into a low-dimensional vector space, enablingthe triples in the knowledge graph to satisfy a specific scoring function in thevector space. However, based on the powerful language understanding and seman-tic modeling capabilities of LLMs, that have recently been invoked to varying degrees in different types of KGE related scenarios such as multi-modal KGE andopen KGE according to their task characteristics. In this paper, we investigate awide range of approaches for performing LLMs-related tasks in different types of KGE scenarios. To better compare the various approaches, we summarize each KGE scenario in a classification. Finally, we discuss the applications in which the methods are mainly used and suggest several forward-looking directions for the development of this new research area.

Bingchen Liu, Huang Peng, Weixin Zeng et al.

The construction of large open knowledge bases (OKBs) is integral to many knowledge-driven applications on the world wide web such as web search. However, noun phrases and relational phrases in OKBs often suffer from redundancy and ambiguity, which calls for the investigation on OKB canonicalization. Current solutions address OKB canonicalization by devising advanced clustering algorithms and using knowledge graph embedding (KGE) to further facilitate the canonicalization process. Nevertheless, these works fail to fully exploit the synergy between clustering and KGE learning, and the methods designed for these subtasks are sub-optimal. To this end, we put forward a multi-task learning framework, namely MulCanon, to tackle OKB canonicalization. In addition, diffusion model is used in the soft clustering process to improve the noun phrase representations with neighboring information, which can lead to more accurate representations. MulCanon unifies the learning objectives of these sub-tasks, and adopts a two-stage multi-task learning paradigm for training. A thorough experimental study on popular OKB canonicalization benchmarks validates that MulCanon can achieve competitive canonicalization results.

Shanchuan Lin, Bingchen Liu, Jiashi Li et al.

We discover that common diffusion noise schedules do not enforce the last timestep to have zero signal-to-noise ratio (SNR), and some implementations of diffusion samplers do not start from the last timestep. Such designs are flawed and do not reflect the fact that the model is given pure Gaussian noise at inference, creating a discrepancy between training and inference. We show that the flawed design causes real problems in existing implementations. In Stable Diffusion, it severely limits the model to only generate images with medium brightness and prevents it from generating very bright and dark samples. We propose a few simple fixes: (1) rescale the noise schedule to enforce zero terminal SNR; (2) train the model with v prediction; (3) change the sampler to always start from the last timestep; (4) rescale classifier-free guidance to prevent over-exposure. These simple changes ensure the diffusion process is congruent between training and inference and allow the model to generate samples more faithful to the original data distribution.

Bingchen Liu, Kunpeng Song, Yizhe Zhu et al.

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.

Bingchen Liu, Kunpeng Song, Ahmed Elgammal

We propose a new approach for synthesizing fully detailed art-stylized images from sketches. Given a sketch, with no semantic tagging, and a reference image of a specific style, the model can synthesize meaningful details with colors and textures. The model consists of three modules designed explicitly for better artistic style capturing and generation. Based on a GAN framework, a dual-masked mechanism is introduced to enforce the content constraints (from the sketch), and a feature-map transformation technique is developed to strengthen the style consistency (to the reference image). Finally, an inverse procedure of instance-normalization is proposed to disentangle the style and content information, therefore yields better synthesis performance. Experiments demonstrate a significant qualitative and quantitative boost over baselines based on previous state-of-the-art techniques, adopted for the proposed process.

Bingchen Liu, Yizhe Zhu, Zuohui Fu et al.

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.

Yizhe Zhu, Jianwen Xie, Bingchen Liu et al.

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.

Ahmed Elgammal, Marian Mazzone, Bingchen Liu et al.

How does the machine classify styles in art? And how does it relate to art historians' methods for analyzing style? Several studies have shown the ability of the machine to learn and predict style categories, such as Renaissance, Baroque, Impressionism, etc., from images of paintings. This implies that the machine can learn an internal representation encoding discriminative features through its visual analysis. However, such a representation is not necessarily interpretable. We conducted a comprehensive study of several of the state-of-the-art convolutional neural networks applied to the task of style classification on 77K images of paintings, and analyzed the learned representation through correlation analysis with concepts derived from art history. Surprisingly, the networks could place the works of art in a smooth temporal arrangement mainly based on learning style labels, without any a priori knowledge of time of creation, the historical time and context of styles, or relations between styles. The learned representations showed that there are few underlying factors that explain the visual variations of style in art. Some of these factors were found to correlate with style patterns suggested by Heinrich Wölfflin (1846-1945). The learned representations also consistently highlighted certain artists as the extreme distinctive representative of their styles, which quantitatively confirms art historian observations.

Yizhe Zhu, Mohamed Elhoseiny, Bingchen Liu et al.

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.

Ahmed Elgammal, Bingchen Liu, Mohamed Elhoseiny et al.

We propose a new system for generating art. The system generates art by looking at art and learning about style; and becomes creative by increasing the arousal potential of the generated art by deviating from the learned styles. We build over Generative Adversarial Networks (GAN), which have shown the ability to learn to generate novel images simulating a given distribution. We argue that such networks are limited in their ability to generate creative products in their original design. We propose modifications to its objective to make it capable of generating creative art by maximizing deviation from established styles and minimizing deviation from art distribution. We conducted experiments to compare the response of human subjects to the generated art with their response to art created by artists. The results show that human subjects could not distinguish art generated by the proposed system from art generated by contemporary artists and shown in top art fairs. Human subjects even rated the generated images higher on various scales.