ACM MM 2024 论文列表

This paper investigates the challenging problem of learned image compression (LIC) with extreme low bitrates. Previous LIC methods based on transmitting quantized continuous features often yield blurry and noisy reconstruction due to the severe quantization loss. While previous LIC methods based on learned codebooks that discretize visual space usually give poor-fidelity reconstruction due to the insufficient representation power of limited codewords in capturing faithful details. We propose a novel dual-stream framework, HyrbidFlow, which combines the continuous-feature-based and codebook-based streams to achieve both high perceptual quality and high fidelity under extreme low bitrates. The codebook-based stream benefits from the high-quality learned codebook priors to provide high quality and clarity in reconstructed images. The continuous feature stream targets at maintaining fidelity details. To achieve the ultra low bitrate, a masked token-based transformer is further proposed, where we only transmit a masked portion of codeword indices and recover the missing indices through token generation guided by information from the continuous feature stream. We also develop a bridging correction network to merge the two streams in pixel decoding for final image reconstruction, where the continuous stream features rectify biases of the codebook-based pixel decoder to impose reconstructed fidelity details. Experimental results demonstrate superior performance across several datasets under extremely low bitrates, compared with existing single-stream codebook-based or continuous-feature-based LIC methods.

Recently, the multi-modal fusion of RGB, depth, and semantics has shown great potential in the domain of dense Simultaneous Localization and Mapping (SLAM), as known as dense semantic SLAM. Yet a prerequisite for generating consistent and continuous semantic maps is the availability of dense, efficient, and scalable scene representations. To date, existing semantic SLAM systems based on explicit scene representations (points/meshes/surfels) are limited by their resolutions and inabilities to predict unknown areas, thus failing to generate dense maps. Contrarily, a few implicit scene representations (Neural Radiance Fields) to deal with these problems rely on time-consuming ray tracing-based volume rendering technique, which cannot meet the real-time rendering requirements of SLAM. Fortunately, the Gaussian Splatting scene representation has recently emerged, which inherits the efficiency and scalability of point/surfel representations while smoothly represents geometric structures in a continuous manner, showing promise in addressing the aforementioned challenges. To this end, we propose GS3LAM, a Gaussian Semantic Splatting SLAM framework, which takes multimodal data as input and can render consistent, continuous dense semantic maps in real-time. To fuse multimodal data, GS3LAM models the scene as a Semantic Gaussian Field (SG-Field), and jointly optimizes camera poses and the field by establishing error constraints between observed and predicted data. Furthermore, a Depth-adaptive Scale Regularization (DSR) scheme is proposed to tackle the problem of misalignment between scale-invariant Gaussians and geometric surfaces within the SG-Field. To mitigate the forgetting phenomenon, we propose an effective Random Sampling-based Keyframe Mapping (RSKM) strategy, which exhibits notable superiority over local covisibility optimization strategies commonly utilized in 3DGS-based SLAM systems. Extensive experiments conducted on the benchmark datasets reveal that compared with state-of-the-art competitors, GS3 LAM demonstrates increased tracking robustness, superior real-time rendering quality, and enhanced semantic reconstruction precision. To make the results reproducible, the source code is available at https://github.com/lif314/GS3LAM.

The rapid progress in generative models has given rise to the critical task of AI-Generated Content Stealth (AIGC-S), which aims to create AI-generated images that can evade both forensic detectors and human inspection. This task is crucial for understanding the vulnerabilities of existing detection methods and developing more robust techniques. However, current adversarial attacks often introduce visible noise, have poor transferability, and fail to address spectral differences between AI-generated and genuine images. To address this, we propose StealthDiffusion, a framework based on stable diffusion that modifies AI-generated images into high-quality, imperceptible adversarial examples capable of evading state-of-the-art forensic detectors. StealthDiffusion comprises two main components: Latent Adversarial Optimization, which generates adversarial perturbations in the latent space of stable diffusion, and Control-VAE, a module that reduces spectral differences between the generated adversarial images and genuine images without affecting the original diffusion model's generation process. Extensive experiments show that StealthDiffusion is effective in both white-box and black-box settings, transforming AI-generated images into high-quality adversarial forgeries with frequency spectra similar to genuine images. These forgeries are classified as genuine by advanced forensic classifiers and are difficult for humans to distinguish.

To address data heterogeneity, the key strategy of Personalized Federated Learning (PFL) is to decouple general knowledge (shared among clients) and client-specific knowledge, as the latter can have a negative impact on collaboration if not removed. Existing PFL methods primarily adopt a parameter partitioning approach, where the parameters of a model are designated as one of two types: parameters shared with other clients to extract general knowledge and parameters retained locally to learn client-specific knowledge. However, as these two types of parameters are put together like a jigsaw puzzle into a single model during the training process, each parameter may simultaneously absorb both general and client-specific knowledge, thus struggling to separate the two types of knowledge effectively. In this paper, we introduce FedDecomp, a simple but effective PFL paradigm that employs parameter additive decomposition to address this issue. Instead of assigning each parameter of a model as either a shared or personalized one, FedDecomp decomposes each parameter into the sum of two parameters: a shared one and a personalized one, thus achieving a more thorough decoupling of shared and personalized knowledge compared to the parameter partitioning method. In addition, as we find that retaining local knowledge of specific clients requires much lower model capacity compared with general knowledge across all clients, we let the matrix containing personalized parameters be low rank during the training process. Moreover, a new alternating training strategy is proposed to further improve the performance. Experimental results across multiple datasets and varying degrees of data heterogeneity demonstrate that FedDecomp outperforms state-of-the-art methods up to 4.9%. The code is available at https://github.com/XinghaoWu/FedDecomp

For solving the limitations of the current self knowledge distillation including never fully utilizing the knowledge of shallow exits and neglecting the impact of auxiliary exits' structure on the performance of network, a novel self knowledge distillation framework via virtual teacher-students mutual learning named LOTH is proposed in this paper. A knowledgeable virtual teacher is constructed from the rich feature maps of each exit to help the learning of each exit. Meanwhile, the logit knowledges of each exit are incorporated to guide the learning of the virtual teacher. They learn mutually through the well-designed loss in LOTH. Moreover, two kinds of auxiliary building blocks are designed to balance the efficiency and effectiveness of network. Extensive experiments with diverse backbones on CIFAR-100 and Tiny-ImageNet validate the effectiveness of LOTH, which realizes superior performance with less resource by the comparison with the state-of-the-art distillation methods. The code of LOTH is available on Github https://github.com/cloak-s/LOTH.

Generating customized content in videos has received increasing attention recently. However, existing works primarily focus on customized text-to-video generation for single subject, suffering from subject-missing and attribute-binding problems when the video is expected to contain multiple subjects. Furthermore, existing models struggle to assign the desired actions to the corresponding subjects (action-binding problem), failing to achieve satisfactory multi-subject generation performance. To tackle the problems, in this paper, we propose DisenStudio, a novel framework that can generate text-guided videos for customized multiple subjects, given few images for each subject. Specifically, DisenStudio enhances a pretrained diffusion-based text-to-video model with our proposed spatial-disentangled cross-attention mechanism to associate each subject with the desired action. Then the model is customized for the multiple subjects with the proposed motion-preserved disentangled finetuning, which involves three tuning strategies: multi-subject co-occurrence tuning, masked single-subject tuning, and multi-subject motion-preserved tuning. The first two strategies guarantee the subject occurrence and preserve their visual attributes, and the third strategy helps the model maintain the temporal motion-generation ability when finetuning on static images. We conduct extensive experiments to demonstrate our proposed DisenStudio significantly outperforms existing methods in various metrics. Additionally, we show that DisenStudio can be used as a powerful tool for various controllable generation applications.

Neural networks often tend to rely on bias features that have strong but spurious correlations with the target labels for decision-making, leading to poor performance on data that does not adhere to these correlations. Early debiasing methods typically construct an unbiased optimization objective based on the labels of bias features. Recent work assumes that bias label is unavailable and usually trains two models: a biased model to deliberately learn bias features for exposing data bias, and a target model to eliminate bias captured by the bias model. In this paper, we first reveal that previous biased models fit target labels, which resulted in failing to expose data bias. To tackle this issue, we propose poisoner, which utilizes data poisoning to embed the biases learned by biased models into the poisoned training data, thereby encouraging the models to learn more biases. Specifically, we couple data poisoning and model training to continuously prompt the biased model to learn more bias. By utilizing the biased model, we can identify samples in the data that contradict these biased correlations. Subsequently, we amplify the influence of these samples in the training of the target model to prevent the model from learning such biased correlations. Experiments show the superior debiasing performance of our method.

A key object in eXplainable Artificial Intelligence (XAI) is to create intelligent systems capable of reasoning and explaining real-world data to facilitate reliable decision-making. Recent studies have acknowledged the importance of providing user-friendly and verifiable explanations to facilitate trustworthy Visual Question Answering (VQA) systems. This paper aims to promote explainable VQA from both data and method perspectives. First, we propose a new Standard Multimodal Explanation (SME) dataset and a new Few-Shot Multimodal Explanation for VQA (FS-MEVQA) task, which aims to generate the multimodal explanation of the underlying reasoning process for solving visual questions with few training samples. Our SME dataset includes 1,028,230 samples composed of questions, images, answers, and multimodal explanations, which can facilitate research in both traditional MEVQA and FS-MEVQA. To the best of our knowledge, this is the first large-scale dataset with joint language-vision explanations based on standard English and additional visual grounding tokens. Second, we propose a training-free Multimodal Explaining Agent (MEAgent) method based on an LLM agent with multimodal open-world tools to infer answers and generate multimodal explanations for visual questions. Our MEAgent can learn multimodal explanation from merely N(=16) training samples and leverage open-world abilities to perform FS-MEVQA on test samples. Comprehensive experimental results evaluated by language quality metrics, visual detection metric, and visual attribution metrics on our SME dataset indicate the superiority of our method for FS-MEVQA. Our code and data are available at https://github.com/LivXue/FS-MEVQA.

Vision-Language Tracking (VLT) requires locating a specific target in video sequences, given a natural language prompt and an initial object box. Despite recent advancements, existing approaches heavily rely on expensive and time-consuming human annotations. To mitigate this limitation, directly generating pseudo labels from raw videos seems to be a straightforward solution; however, it inevitably introduces undesirable noise during the training process. Moreover, we insist that an efficient tracker should excel in tracking the target, regardless of the temporal direction. Building upon these insights, we propose the pioneering semi-supervised learning scheme for VLT task, representing a crucial step towards reducing the dependency on high-quality yet costly labeled data. Specifically, drawing inspiration from the natural attributes of a video (i.e., space, time, and semantics), our approach progressively leverages inherent consistencies from these aspects: (1) Spatially, each frame and any object cropped from it naturally form an image-bbox (bounding box) pair for self-training; (2) Temporally, bidirectional tracking trajectories should exhibit minimal differences; (3) Semantically, the correlation between visual and textual features is expected to remain consistent. Furthermore, the framework is validated with a simple yet effective tracker we devised, named ATTracker (Asymmetrical Transformer Tracker). It modifies the self-attention operation in an asymmetrical way, striving to enhance target-related features while suppressing noise. Extensive experiments confirm that our ATTracker serves as a robust baseline, outperforming fully supervised base trackers. By unveiling the potential of learning with limited annotations, this study aims to attract attention and pave the way for Semi-supervised Vision-Language Tracking (SS-VLT).

Miscalibrated models tend to be unreliable and insecure for downstream applications. In this work, we attempt to highlight and remedy miscalibration in current scene graph generation (SGG) models, which has been overlooked by previous works. We discover that obtaining well-calibrated models for SGG is more challenging than conventional calibration settings, as long-tailed SGG training data exacerbates miscalibration with overconfidence in head classes and underconfidence in tail classes. We further analyze which components are explicitly impacted by the long-tailed data during optimization, thereby exacerbating miscalibration and unbalanced learning, including biased parameters, deviated boundaries, and distorted target distribution. To address the above issues, we propose the Compositional Optimization Calibration (COC) method, comprising three modules: i. A parameter calibration module that utilizes a hyperspherical classifier to eliminate the bias introduced by biased parameters. ii. A boundary calibration module that disperses features of majority classes to consolidate the decision boundaries of minority classes and mitigate deviated boundaries. iii. A target distribution calibration module that addresses distorted target distribution, leverages within-triplet prior to guide confidence-aware and label-aware target calibration, and applies curriculum regulation to constrain learning focus from easy to hard classes. Extensive evaluation on popular benchmarks demonstrates the effectiveness of our proposed method in improving model calibration and resolving unbalanced learning for long-tailed SGG. Finally, our proposed method performs best on model calibration compared to different types of calibration methods and achieves state-of-the-art trade-off performance on balanced SGG learning.

Inserting foreground objects into specific background scenes and eliminating the illumination inconsistency (eg., color, brightness) between them is an important and challenging task. It typically involves multiple processing tasks, such as image harmonization and shadow generation. In these two domains, there are already many mature solutions, but they often only focus on one of the tasks. Recently, some image composition methods have utilized diffusion models to address both of these issues simultaneously, but they cannot guarantee complete reconstruction of the foreground content. In this work, we propose CFDiffusion, which can simultaneously handle image harmonization and shadow generation. We first employ a shadow mask predictor to estimate the shadow mask of the foreground object. Next, we design a harmonization-shadow generator based on a diffusion model to harmonize the foreground and generate shadows concurrently. Additionally, we propose a foreground content enhancement module to ensure the complete preservation of foreground content at the insertion location, and we also develop an adaptive encoder to guide the harmonization process in the foreground area. The experimental results on the iHarmony4 dataset and the IH-SG dataset demonstrate the superiority of our CFDiffusion approach.

Images corrupted by rain streaks often lose vital frequency information for perception, and image deraining aims to solve this problem, which relies on global and local degradation modeling. Recent studies have witnessed the effectiveness and efficiency of Mamba for perceiving global and local information based on its exploiting local correlation among patches, however, rarely attempts have been explored to extend it with frequency analysis for image deraining, limiting its ability to perceive global degradation that is relevant to frequency modeling (e.g. Fourier transform). In this paper, we propose FreqMamba, an effective and efficient paradigm that leverages the complementary between Mamba and frequency analysis for image deraining. The core of our method lies in extending Mamba with frequency analysis from two perspectives: extending it with frequency band for exploiting frequency correlation, and connecting it with Fourier transform for global degradation modeling. Specifically, FreqMamba introduces complementary triple interaction structures including spatial Mamba, frequency-band Mamba, and Fourier global modeling. Frequency-Band Mamba decomposes the image into sub-bands of different frequencies to allow 2D scanning from the frequency dimension. Furthermore, leveraging Mamba's unique data-dependent properties, we use rainy images at different scales to provide degradation priors to the network, thereby facilitating efficient training. Extensive experiments show that our method outperforms state-of-the-art methods both visually and quantitatively. Our code is available at: https://github.com/aSleepyTree/FreqMamba.

Existing RGB-D semantic segmentation methods struggle to handle modality missing input, where only RGB images or depth maps are available, leading to degenerated segmentation performance. We tackle this issue using MaskMentor, a new pre-training framework for modality missing segmentation, which advances its counterparts via two novel designs: Masked Modality and Image Modeling (M2IM), and Self-Teaching via Token-Pixel Joint reconstruction (STTP). M2IM simulates modality missing scenarios by combining both modality- and patch-level random masking. Meanwhile, STTP offers an effective self-teaching strategy, where the trained network assumes a dual role, simultaneously acting as both the teacher and the student. The student with modality missing input is supervised by the teacher with complete modality input through both token- and pixel-wise masked modeling, closing the gap between missing and complete input modalities. By integrating M2IM and STTP, MaskMentor significantly improves the generalization ability of the trained model across diverse input conditions and outperforms state-of-the-art methods on two popular benchmarks by a considerable margin. Extensive ablation studies further verify the effectiveness of the above contributions.

Automatically narrating videos in natural language complying with user requests, i.e. Controllable Video Captioning task, can help people manage massive videos with desired intentions. However, existing works suffer from two shortcomings: 1) the control signal is single-grained which can not satisfy diverse user intentions; 2) the video description is generated in a single round which can not be further edited to meet dynamic needs. In this paper, we propose a novel Video Caption Editing (VCE) task to automatically revise an existing video description guided by multi-grained user requests. Inspired by human writing-revision habits, we design the user command as a pivotal triplet {operation, position, attribute} to cover diverse user needs from coarse-grained to fine-grained. To facilitate the VCE task, we automatically construct an open-domain benchmark dataset named VATEX-EDIT and manually collect an e-commerce dataset called EMMAD-EDIT. We further propose a specialized small-scale model (i.e., OPA) compared with two generalist Large Multi-modal Models to perform an exhaustive analysis of the novel task. For evaluation, we adopt comprehensive metrics considering caption fluency, command-caption consistency, and video-caption alignment. Experiments reveal the task challenges of fine-grained multi-modal semantics understanding and processing. Our datasets, codes, and evaluation tools are available at https://github.com/yaolinli/VCE.

Sparse and noisy images (SNIs), like those in spatial gene expression data, pose significant challenges for effective representation learning and clustering, which are essential for thorough data analysis and interpretation. In response to these challenges, we propose Dual Advancement of Representation Learning and Clustering (DARLC), an innovative framework that leverages contrastive learning to enhance the representations derived from masked image modeling. Simultaneously, DARLC integrates cluster assignments in a cohesive, end-to-end approach. This integrated clustering strategy addresses the "class collision problem" inherent in contrastive learning, thus improving the quality of the resulting representations. To generate more plausible positive views for contrastive learning, we employ a graph attention network-based technique that produces denoised images as augmented data. As such, our framework offers a comprehensive approach that improves the learning of representations by enhancing their local perceptibility, distinctiveness, and the understanding of relational semantics. Furthermore, we utilize a Student's t mixture model to achieve more robust and adaptable clustering of SNIs. Extensive experiments, conducted across 12 different types of datasets consisting of SNIs, demonstrate that DARLC surpasses the state-of-the-art methods in both image clustering and generating image representations that accurately capture gene interactions. Code is available at https://github.com/zipging/DARLC.