Enhancing Image Generation Fidelity via Progressive Prompts

The diffusion transformer (DiT) architecture has attracted significant attention in image generation, achieving better fidelity, performance, and diversity. However, most existing DiT - based image generation methods focus on global - aware synthesis, and regional prompt control has been less explored. In this paper, we propose a coarse - to - fine generation pipeline for regional prompt - following generation. Specifically, we first utilize the powerful large language model (LLM) to generate both high - level descriptions of the image (such as content, topic, and objects) and low - level descriptions (such as details and style). Then, we explore the influence of cross - attention layers at different depths. We find that deeper layers are always responsible for high - level content control, while shallow layers handle low - level content control. Various prompts are injected into the proposed regional cross - attention control for coarse - to - fine generation. By using the proposed pipeline, we enhance the controllability of DiT - based image generation. Extensive quantitative and qualitative results show that our pipeline can improve the performance of the generated images.

VASparse: Towards Efficient Visual Hallucination Mitigation for Large Vision-Language Model via Visual-Aware Sparsification

Large Vision-Language Models (LVLMs) may produce outputs that are unfaithful to reality, also known as visual hallucinations (VH), which significantly impedes their real-world usage. To alleviate VH, various decoding strategies have been proposed to enhance visual information. However, many of these methods may require secondary decoding and rollback, which significantly reduces inference speed. In this work, we propose an efficient plug-and-play decoding algorithm via Visual-Aware Sparsification (VASparse) from the perspective of token sparsity for mitigating VH. VASparse is inspired by empirical observations: (1) the sparse activation of attention in LVLMs, and (2) visual-agnostic tokens sparsification exacerbates VH. Based on these insights, we propose a novel token sparsification strategy that balances efficiency and trustworthiness. Specifically, VASparse implements a visual-aware token selection strategy during decoding to reduce redundant tokens while preserving visual context effectively. Additionally, we innovatively introduce a sparse-based visual contrastive decoding method to recalibrate the distribution of hallucinated outputs without the time overhead associated with secondary decoding. Subsequently, VASparse recalibrates attention scores to penalize attention sinking of LVLMs towards text tokens. Extensive experiments across four popular benchmarks confirm the effectiveness of VASparse in mitigating VH across different LVLM families without requiring additional training or post-processing. Impressively, VASparse achieves state-of-the-art performance for mitigating VH while maintaining competitive decoding speed. Code is available at https://github.com/mengchuang123/VASparse-github.

DisPose: Disentangling Pose Guidance for Controllable Human Image Animation

Controllable human image animation aims to generate videos from reference images using driving videos. Due to the limited control signals provided by sparse guidance (e.g., skeleton pose), recent works have attempted to introduce additional dense conditions (e.g., depth map) to ensure motion alignment. However, such strict dense guidance impairs the quality of the generated video when the body shape of the reference character differs significantly from that of the driving video. In this paper, we present DisPose to mine more generalizable and effective control signals without additional dense input, which disentangles the sparse skeleton pose in human image animation into motion field guidance and keypoint correspondence. Specifically, we generate a dense motion field from a sparse motion field and the reference image, which provides region-level dense guidance while maintaining the generalization of the sparse pose control. We also extract diffusion features corresponding to pose keypoints from the reference image, and then these point features are transferred to the target pose to provide distinct identity information. To seamlessly integrate into existing models, we propose a plug-and-play hybrid ControlNet that improves the quality and consistency of generated videos while freezing the existing model parameters. Extensive qualitative and quantitative experiments demonstrate the superiority of DisPose compared to current methods. Code: \href{https://github.com/lihxxx/DisPose}{https://github.com/lihxxx/DisPose}.

DiffATR: Diffusion-based Generative Modeling for Audio-Text Retrieval

Existing audio-text retrieval (ATR) methods are essentially discriminative models that aim to maximize the conditional likelihood, represented as p(candidates|query). Nevertheless, this methodology fails to consider the intrinsic data distribution p(query), leading to difficulties in discerning out-of-distribution data. In this work, we attempt to tackle this constraint through a generative perspective and model the relationship between audio and text as their joint probability p(candidates,query). To this end, we present a diffusion-based ATR framework (DiffATR), which models ATR as an iterative procedure that progressively generates joint distribution from noise. Throughout its training phase, DiffATR is optimized from both generative and discriminative viewpoints: the generator is refined through a generation loss, while the feature extractor benefits from a contrastive loss, thus combining the merits of both methodologies. Experiments on the AudioCaps and Clotho datasets with superior performances, verify the effectiveness of our approach. Notably, without any alterations, our DiffATR consistently exhibits strong performance in out-of-domain retrieval settings.

Audio-text Retrieval with Transformer-based Hierarchical Alignment and Disentangled Cross-modal Representation

Most existing audio-text retrieval (ATR) approaches typically rely on a single-level interaction to associate audio and text, limiting their ability to align different modalities and leading to suboptimal matches. In this work, we present a novel ATR framework that leverages two-stream Transformers in conjunction with a Hierarchical Alignment (THA) module to identify multi-level correspondences of different Transformer blocks between audio and text. Moreover, current ATR methods mainly focus on learning a global-level representation, missing out on intricate details to capture audio occurrences that correspond to textual semantics. To bridge this gap, we introduce a Disentangled Cross-modal Representation (DCR) approach that disentangles high-dimensional features into compact latent factors to grasp fine-grained audio-text semantic correlations. Additionally, we develop a confidence-aware (CA) module to estimate the confidence of each latent factor pair and adaptively aggregate cross-modal latent factors to achieve local semantic alignment. Experiments show that our THA effectively boosts ATR performance, with the DCR approach further contributing to consistent performance gains.

Recent Trends of Multimodal Affective Computing: A Survey from NLP Perspective

Multimodal affective computing (MAC) has garnered increasing attention due to its broad applications in analyzing human behaviors and intentions, especially in text-dominated multimodal affective computing field. This survey presents the recent trends of multimodal affective computing from NLP perspective through four hot tasks: multimodal sentiment analysis, multimodal emotion recognition in conversation, multimodal aspect-based sentiment analysis and multimodal multi-label emotion recognition. The goal of this survey is to explore the current landscape of multimodal affective research, identify development trends, and highlight the similarities and differences across various tasks, offering a comprehensive report on the recent progress in multimodal affective computing from an NLP perspective. This survey covers the formalization of tasks, provides an overview of relevant works, describes benchmark datasets, and details the evaluation metrics for each task. Additionally, it briefly discusses research in multimodal affective computing involving facial expressions, acoustic signals, physiological signals, and emotion causes. Additionally, we discuss the technical approaches, challenges, and future directions in multimodal affective computing. To support further research, we released a repository that compiles related works in multimodal affective computing, providing detailed resources and references for the community.

XMeCap: Meme Caption Generation with Sub-Image Adaptability

Humor, deeply rooted in societal meanings and cultural details, poses a unique challenge for machines. While advances have been made in natural language processing, real-world humor often thrives in a multi-modal context, encapsulated distinctively by memes. This paper poses a particular emphasis on the impact of multi-images on meme captioning. After that, we introduce the \textsc{XMeCap} framework, a novel approach that adopts supervised fine-tuning and reinforcement learning based on an innovative reward model, which factors in both global and local similarities between visuals and text. Our results, benchmarked against contemporary models, manifest a marked improvement in caption generation for both single-image and multi-image memes, as well as different meme categories. \textsc{XMeCap} achieves an average evaluation score of 75.85 for single-image memes and 66.32 for multi-image memes, outperforming the best baseline by 3.71\% and 4.82\%, respectively. This research not only establishes a new frontier in meme-related studies but also underscores the potential of machines in understanding and generating humor in a multi-modal setting.

MJ-Bench: Is Your Multimodal Reward Model Really a Good Judge for Text-to-Image Generation?

While text-to-image models like DALLE-3 and Stable Diffusion are rapidly proliferating, they often encounter challenges such as hallucination, bias, and the production of unsafe, low-quality output. To effectively address these issues, it is crucial to align these models with desired behaviors based on feedback from a multimodal judge. Despite their significance, current multimodal judges frequently undergo inadequate evaluation of their capabilities and limitations, potentially leading to misalignment and unsafe fine-tuning outcomes. To address this issue, we introduce MJ-Bench, a novel benchmark which incorporates a comprehensive preference dataset to evaluate multimodal judges in providing feedback for image generation models across four key perspectives: alignment, safety, image quality, and bias. Specifically, we evaluate a large variety of multimodal judges including smaller-sized CLIP-based scoring models, open-source VLMs (e.g. LLaVA family), and close-source VLMs (e.g. GPT-4o, Claude 3) on each decomposed subcategory of our preference dataset. Experiments reveal that close-source VLMs generally provide better feedback, with GPT-4o outperforming other judges in average. Compared with open-source VLMs, smaller-sized scoring models can provide better feedback regarding text-image alignment and image quality, while VLMs provide more accurate feedback regarding safety and generation bias due to their stronger reasoning capabilities. Further studies in feedback scale reveal that VLM judges can generally provide more accurate and stable feedback in natural language (Likert-scale) than numerical scales. Notably, human evaluations on end-to-end fine-tuned models using separate feedback from these multimodal judges provide similar conclusions, further confirming the effectiveness of MJ-Bench. All data, code, models are available at https://huggingface.co/MJ-Bench.

LOOK-M: Look-Once Optimization in KV Cache for Efficient Multimodal Long-Context Inference

Long-context Multimodal Large Language Models (MLLMs) demand substantial computational resources for inference as the growth of their multimodal Key-Value (KV) cache, in response to increasing input lengths, challenges memory and time efficiency. Unlike single-modality LLMs that manage only textual contexts, the KV cache of long-context MLLMs includes representations from multiple images with temporal and spatial relationships and related textual contexts. The predominance of image tokens means traditional optimizations for LLMs' KV caches are unsuitable for multimodal long-context settings, and no prior works have addressed this challenge. In this work, we introduce LOOK-M, a pioneering, fine-tuning-free approach that efficiently reduces the multimodal KV cache size while maintaining performance comparable to a full cache. We observe that during prompt prefill, the model prioritizes more textual attention over image features, and based on the multimodal interaction observation, a new proposed text-prior method is explored to compress the KV cache. Furthermore, to mitigate the degradation of image contextual information, we propose several compensatory strategies using KV pairs merging. LOOK-M demonstrates that with a significant reduction in KV Cache memory usage, such as reducing it by 80% in some cases, it not only achieves up to 1.5x faster decoding but also maintains or even enhances performance across a variety of long context multimodal tasks.

D2O:Dynamic Discriminative Operations for Efficient Generative Inference of Large Language Models

Efficient inference in Large Language Models (LLMs) is impeded by the growing memory demands of key-value (KV) caching, especially for longer sequences. Traditional KV cache eviction strategies, which prioritize less critical KV-pairs based on attention scores, often degrade generation quality, leading to issues such as context loss or hallucinations. To address this, we introduce Dynamic Discriminative Operations (D2O), a novel method that utilizes two-level discriminative strategies to optimize KV cache size without fine-tuning, while preserving essential context. Initially, by observing varying densities of attention weights between shallow and deep layers, we use this insight to determine which layers should avoid excessive eviction to minimize information loss. Subsequently, for the eviction strategy in each layer, D2O innovatively incorporates a compensation mechanism that maintains a similarity threshold to re-discriminate the importance of previously discarded tokens, determining whether they should be recalled and merged with similar tokens. Our approach not only achieves significant memory savings and enhances inference throughput by more than 3x but also maintains high-quality long-text generation. Extensive experiments across various benchmarks and LLM architectures have demonstrated that D2O significantly enhances performance with a constrained KV cache budget.