Empowering Backbone Models for Visual Text Generation with Input Granularity Control and Glyph-Aware Training

Diffusion-based text-to-image models have demonstrated impressive achievements in diversity and aesthetics but struggle to generate images with legible visual texts. Existing backbone models have limitations such as misspelling, failing to generate texts, and lack of support for Chinese text, but their development shows promising potential. In this paper, we propose a series of methods, aiming to empower backbone models to generate visual texts in English and Chinese. We first conduct a preliminary study revealing that Byte Pair Encoding (BPE) tokenization and the insufficient learning of cross-attention modules restrict the performance of the backbone models. Based on these observations, we make the following improvements: (1) We design a mixed granularity input strategy to provide more suitable text representations; (2) We propose to augment the conventional training objective with three glyph-aware training losses, which enhance the learning of cross-attention modules and encourage the model to focus on visual texts. Through experiments, we demonstrate that our methods can effectively empower backbone models to generate semantic relevant, aesthetically appealing, and accurate visual text images, while maintaining their fundamental image generation quality.

MonoFormer: One Transformer for Both Diffusion and Autoregression

Most existing multimodality methods use separate backbones for autoregression-based discrete text generation and diffusion-based continuous visual generation, or the same backbone by discretizing the visual data to use autoregression for both text and visual generation. In this paper, we propose to study a simple idea: share one transformer for both autoregression and diffusion. The feasibility comes from two main aspects: (i) Transformer is successfully applied to diffusion for visual generation, and (ii) transformer training for autoregression and diffusion is very similar, and the difference merely lies in that diffusion uses bidirectional attention mask and autoregression uses causal attention mask. Experimental results show that our approach achieves comparable image generation performance to current state-of-the-art methods as well as maintains the text generation capability. The project is publicly available at https://monoformer.github.io/.

UNIMO-G: Unified Image Generation through Multimodal Conditional Diffusion

Existing text-to-image diffusion models primarily generate images from text prompts. However, the inherent conciseness of textual descriptions poses challenges in faithfully synthesizing images with intricate details, such as specific entities or scenes. This paper presents UNIMO-G, a simple multimodal conditional diffusion framework that operates on multimodal prompts with interleaved textual and visual inputs, which demonstrates a unified ability for both text-driven and subject-driven image generation. UNIMO-G comprises two core components: a Multimodal Large Language Model (MLLM) for encoding multimodal prompts, and a conditional denoising diffusion network for generating images based on the encoded multimodal input. We leverage a two-stage training strategy to effectively train the framework: firstly pre-training on large-scale text-image pairs to develop conditional image generation capabilities, and then instruction tuning with multimodal prompts to achieve unified image generation proficiency. A well-designed data processing pipeline involving language grounding and image segmentation is employed to construct multi-modal prompts. UNIMO-G excels in both text-to-image generation and zero-shot subject-driven synthesis, and is notably effective in generating high-fidelity images from complex multimodal prompts involving multiple image entities.

UniVG: Towards UNIfied-modal Video Generation

Diffusion based video generation has received extensive attention and achieved considerable success within both the academic and industrial communities. However, current efforts are mainly concentrated on single-objective or single-task video generation, such as generation driven by text, by image, or by a combination of text and image. This cannot fully meet the needs of real-world application scenarios, as users are likely to input images and text conditions in a flexible manner, either individually or in combination. To address this, we propose a Unified-modal Video Genearation system that is capable of handling multiple video generation tasks across text and image modalities. To this end, we revisit the various video generation tasks within our system from the perspective of generative freedom, and classify them into high-freedom and low-freedom video generation categories. For high-freedom video generation, we employ Multi-condition Cross Attention to generate videos that align with the semantics of the input images or text. For low-freedom video generation, we introduce Biased Gaussian Noise to replace the pure random Gaussian Noise, which helps to better preserve the content of the input conditions. Our method achieves the lowest Fr\'echet Video Distance (FVD) on the public academic benchmark MSR-VTT, surpasses the current open-source methods in human evaluations, and is on par with the current close-source method Gen2. For more samples, visit https://univg-baidu.github.io.

HiCAST: Highly Customized Arbitrary Style Transfer with Adapter Enhanced Diffusion Models

The goal of Arbitrary Style Transfer (AST) is injecting the artistic features of a style reference into a given image/video. Existing methods usually focus on pursuing the balance between style and content, whereas ignoring the significant demand for flexible and customized stylization results and thereby limiting their practical application. To address this critical issue, a novel AST approach namely HiCAST is proposed, which is capable of explicitly customizing the stylization results according to various source of semantic clues. In the specific, our model is constructed based on Latent Diffusion Model (LDM) and elaborately designed to absorb content and style instance as conditions of LDM. It is characterized by introducing of \textit{Style Adapter}, which allows user to flexibly manipulate the output results by aligning multi-level style information and intrinsic knowledge in LDM. Lastly, we further extend our model to perform video AST. A novel learning objective is leveraged for video diffusion model training, which significantly improve cross-frame temporal consistency in the premise of maintaining stylization strength. Qualitative and quantitative comparisons as well as comprehensive user studies demonstrate that our HiCAST outperforms the existing SoTA methods in generating visually plausible stylization results.

Exploiting Diffusion Priors for All-in-One Image Restoration

All-in-one aims to solve various tasks of image restoration in a single model. To this end, we present a feasible way of exploiting the image priors captured by the pretrained diffusion model, through addressing the two challenges, i.e., degradation modeling and diffusion guidance. The former aims to simulate the process of the clean image degenerated by certain degradations, and the latter aims at guiding the diffusion model to generate the corresponding clean image. With the motivations, we propose a zero-shot framework for all-in-one image restoration, termed ZeroAIR, which alternatively performs the test-time degradation modeling (TDM) and the three-stage diffusion guidance (TDG) at each timestep of the reverse sampling. To be specific, TDM exploits the diffusion priors to learn a degradation model from a given degraded image, and TDG divides the timesteps into three stages for taking full advantage of the varying diffusion priors. Thanks to their degradation-agnostic property, the all-in-one image restoration could be achieved in a zero-shot way by ZeroAIR. Through extensive experiments, we show that our ZeroAIR achieves comparable even better performance than those task-specific methods. The code will be available on Github.

UNIMO-3: Multi-granularity Interaction for Vision-Language Representation Learning

Vision-and-language (VL) pre-training, which aims to learn a general representation of image-text pairs that can be transferred to various vision-and-language tasks. Compared with modeling uni-modal data, the main challenge of the VL model is: how to learn the cross-modal interaction from multimodal data, especially the fine-grained interaction. Existing works have shown that fully transformer-based models that adopt attention mechanisms to learn in-layer cross-model interaction can demonstrate impressive performance on various cross-modal downstream tasks. However, they ignored that the semantic information of the different modals at the same layer was not uniform, which leads to the cross-modal interaction collapsing into a limited multi-modal semantic information interaction. In this work, we propose the UNIMO-3 model, which has the capacity to simultaneously learn the multimodal in-layer interaction and cross-layer interaction. UNIMO-3 model can establish effective connections between different layers in a cross-modal encoder, and adaptively capture the interaction between two modalities at different levels. The experimental results show that our model achieves state-of-the-art performance in various downstream tasks, and through ablation study can prove that effective cross-layer learning improves the model's ability of multimodal representation.

WeCheck: Strong Factual Consistency Checker via Weakly Supervised Learning

A crucial issue of current text generation models is that they often uncontrollably generate factually inconsistent text with respective of their inputs. Limited by the lack of annotated data, existing works in evaluating factual consistency directly transfer the reasoning ability of models trained on other data-rich upstream tasks like question answering (QA) and natural language inference (NLI) without any further adaptation. As a result, they perform poorly on the real generated text and are biased heavily by their single-source upstream tasks. To alleviate this problem, we propose a weakly supervised framework that aggregates multiple resources to train a precise and efficient factual metric, namely WeCheck. WeCheck first utilizes a generative model to accurately label a real generated sample by aggregating its weak labels, which are inferred from multiple resources. Then, we train the target metric model with the weak supervision while taking noises into consideration. Comprehensive experiments on a variety of tasks demonstrate the strong performance of WeCheck, which achieves a 3.4\% absolute improvement over previous state-of-the-art methods on TRUE benchmark on average.

UPainting: Unified Text-to-Image Diffusion Generation with Cross-modal Guidance

Diffusion generative models have recently greatly improved the power of text-conditioned image generation. Existing image generation models mainly include text conditional diffusion model and cross-modal guided diffusion model, which are good at small scene image generation and complex scene image generation respectively. In this work, we propose a simple yet effective approach, namely UPainting, to unify simple and complex scene image generation, as shown in Figure 1. Based on architecture improvements and diverse guidance schedules, UPainting effectively integrates cross-modal guidance from a pretrained image-text matching model into a text conditional diffusion model that utilizes a pretrained Transformer language model as the text encoder. Our key findings is that combining the power of large-scale Transformer language model in understanding language and image-text matching model in capturing cross-modal semantics and style, is effective to improve sample fidelity and image-text alignment of image generation. In this way, UPainting has a more general image generation capability, which can generate images of both simple and complex scenes more effectively. To comprehensively compare text-to-image models, we further create a more general benchmark, UniBench, with well-written Chinese and English prompts in both simple and complex scenes. We compare UPainting with recent models and find that UPainting greatly outperforms other models in terms of caption similarity and image fidelity in both simple and complex scenes. UPainting project page \url{https://upainting.github.io/}.

FRSUM: Towards Faithful Abstractive Summarization via Enhancing Factual Robustness

Despite being able to generate fluent and grammatical text, current Seq2Seq summarization models still suffering from the unfaithful generation problem. In this paper, we study the faithfulness of existing systems from a new perspective of factual robustness which is the ability to correctly generate factual information over adversarial unfaithful information. We first measure a model's factual robustness by its success rate to defend against adversarial attacks when generating factual information. The factual robustness analysis on a wide range of current systems shows its good consistency with human judgments on faithfulness. Inspired by these findings, we propose to improve the faithfulness of a model by enhancing its factual robustness. Specifically, we propose a novel training strategy, namely FRSUM, which teaches the model to defend against both explicit adversarial samples and implicit factual adversarial perturbations. Extensive automatic and human evaluation results show that FRSUM consistently improves the faithfulness of various Seq2Seq models, such as T5, BART.