EmoDubber: Towards High Quality and Emotion Controllable Movie Dubbing

Given a piece of text, a video clip, and a reference audio, the movie dubbing task aims to generate speech that aligns with the video while cloning the desired voice. The existing methods have two primary deficiencies: (1) They struggle to simultaneously hold audio-visual sync and achieve clear pronunciation; (2) They lack the capacity to express user-defined emotions. To address these problems, we propose EmoDubber, an emotion-controllable dubbing architecture that allows users to specify emotion type and emotional intensity while satisfying high-quality lip sync and pronunciation. Specifically, we first design Lip-related Prosody Aligning (LPA), which focuses on learning the inherent consistency between lip motion and prosody variation by duration level contrastive learning to incorporate reasonable alignment. Then, we design Pronunciation Enhancing (PE) strategy to fuse the video-level phoneme sequences by efficient conformer to improve speech intelligibility. Next, the speaker identity adapting module aims to decode acoustics prior and inject the speaker style embedding. After that, the proposed Flow-based User Emotion Controlling (FUEC) is used to synthesize waveform by flow matching prediction network conditioned on acoustics prior. In this process, the FUEC determines the gradient direction and guidance scale based on the user's emotion instructions by the positive and negative guidance mechanism, which focuses on amplifying the desired emotion while suppressing others. Extensive experimental results on three benchmark datasets demonstrate favorable performance compared to several state-of-the-art methods.

Expanding Sparse Tuning for Low Memory Usage

Parameter-efficient fine-tuning (PEFT) is an effective method for adapting pre-trained vision models to downstream tasks by tuning a small subset of parameters. Among PEFT methods, sparse tuning achieves superior performance by only adjusting the weights most relevant to downstream tasks, rather than densely tuning the whole weight matrix. However, this performance improvement has been accompanied by increases in memory usage, which stems from two factors, i.e., the storage of the whole weight matrix as learnable parameters in the optimizer and the additional storage of tunable weight indexes. In this paper, we propose a method named SNELL (Sparse tuning with kerNELized LoRA) for sparse tuning with low memory usage. To achieve low memory usage, SNELL decomposes the tunable matrix for sparsification into two learnable low-rank matrices, saving from the costly storage of the whole original matrix. A competition-based sparsification mechanism is further proposed to avoid the storage of tunable weight indexes. To maintain the effectiveness of sparse tuning with low-rank matrices, we extend the low-rank decomposition by applying nonlinear kernel functions to the whole-matrix merging. Consequently, we gain an increase in the rank of the merged matrix, enhancing the ability of SNELL in adapting the pre-trained models to downstream tasks. Extensive experiments on multiple downstream tasks show that SNELL achieves state-of-the-art performance with low memory usage, endowing PEFT with sparse tuning to large-scale models. Codes are available at https://github.com/ssfgunner/SNELL.

Towards Dynamic Message Passing on Graphs

Message passing plays a vital role in graph neural networks (GNNs) for effective feature learning. However, the over-reliance on input topology diminishes the efficacy of message passing and restricts the ability of GNNs. Despite efforts to mitigate the reliance, existing study encounters message-passing bottlenecks or high computational expense problems, which invokes the demands for flexible message passing with low complexity. In this paper, we propose a novel dynamic message-passing mechanism for GNNs. It projects graph nodes and learnable pseudo nodes into a common space with measurable spatial relations between them. With nodes moving in the space, their evolving relations facilitate flexible pathway construction for a dynamic message-passing process. Associating pseudo nodes to input graphs with their measured relations, graph nodes can communicate with each other intermediately through pseudo nodes under linear complexity. We further develop a GNN model named $\mathtt{\mathbf{N^2}}$ based on our dynamic message-passing mechanism. $\mathtt{\mathbf{N^2}}$ employs a single recurrent layer to recursively generate the displacements of nodes and construct optimal dynamic pathways. Evaluation on eighteen benchmarks demonstrates the superior performance of $\mathtt{\mathbf{N^2}}$ over popular GNNs. $\mathtt{\mathbf{N^2}}$ successfully scales to large-scale benchmarks and requires significantly fewer parameters for graph classification with the shared recurrent layer.

Prompting Video-Language Foundation Models with Domain-specific Fine-grained Heuristics for Video Question Answering

Video Question Answering (VideoQA) represents a crucial intersection between video understanding and language processing, requiring both discriminative unimodal comprehension and sophisticated cross-modal interaction for accurate inference. Despite advancements in multi-modal pre-trained models and video-language foundation models, these systems often struggle with domain-specific VideoQA due to their generalized pre-training objectives. Addressing this gap necessitates bridging the divide between broad cross-modal knowledge and the specific inference demands of VideoQA tasks. To this end, we introduce HeurVidQA, a framework that leverages domain-specific entity-action heuristics to refine pre-trained video-language foundation models. Our approach treats these models as implicit knowledge engines, employing domain-specific entity-action prompters to direct the model's focus toward precise cues that enhance reasoning. By delivering fine-grained heuristics, we improve the model's ability to identify and interpret key entities and actions, thereby enhancing its reasoning capabilities. Extensive evaluations across multiple VideoQA datasets demonstrate that our method significantly outperforms existing models, underscoring the importance of integrating domain-specific knowledge into video-language models for more accurate and context-aware VideoQA.

Multi-granularity Contrastive Cross-modal Collaborative Generation for End-to-End Long-term Video Question Answering

Long-term Video Question Answering (VideoQA) is a challenging vision-and-language bridging task focusing on semantic understanding of untrimmed long-term videos and diverse free-form questions, simultaneously emphasizing comprehensive cross-modal reasoning to yield precise answers. The canonical approaches often rely on off-the-shelf feature extractors to detour the expensive computation overhead, but often result in domain-independent modality-unrelated representations. Furthermore, the inherent gradient blocking between unimodal comprehension and cross-modal interaction hinders reliable answer generation. In contrast, recent emerging successful video-language pre-training models enable cost-effective end-to-end modeling but fall short in domain-specific ratiocination and exhibit disparities in task formulation. Toward this end, we present an entirely end-to-end solution for long-term VideoQA: Multi-granularity Contrastive cross-modal collaborative Generation (MCG) model. To derive discriminative representations possessing high visual concepts, we introduce Joint Unimodal Modeling (JUM) on a clip-bone architecture and leverage Multi-granularity Contrastive Learning (MCL) to harness the intrinsically or explicitly exhibited semantic correspondences. To alleviate the task formulation discrepancy problem, we propose a Cross-modal Collaborative Generation (CCG) module to reformulate VideoQA as a generative task instead of the conventional classification scheme, empowering the model with the capability for cross-modal high-semantic fusion and generation so as to rationalize and answer. Extensive experiments conducted on six publicly available VideoQA datasets underscore the superiority of our proposed method.

Suppress Content Shift: Better Diffusion Features via Off-the-Shelf Generation Techniques

Diffusion models are powerful generative models, and this capability can also be applied to discrimination. The inner activations of a pre-trained diffusion model can serve as features for discriminative tasks, namely, diffusion feature. We discover that diffusion feature has been hindered by a hidden yet universal phenomenon that we call content shift. To be specific, there are content differences between features and the input image, such as the exact shape of a certain object. We locate the cause of content shift as one inherent characteristic of diffusion models, which suggests the broad existence of this phenomenon in diffusion feature. Further empirical study also indicates that its negative impact is not negligible even when content shift is not visually perceivable. Hence, we propose to suppress content shift to enhance the overall quality of diffusion features. Specifically, content shift is related to the information drift during the process of recovering an image from the noisy input, pointing out the possibility of turning off-the-shelf generation techniques into tools for content shift suppression. We further propose a practical guideline named GATE to efficiently evaluate the potential benefit of a technique and provide an implementation of our methodology. Despite the simplicity, the proposed approach has achieved superior results on various tasks and datasets, validating its potential as a generic booster for diffusion features. Our code is available at https://github.com/Darkbblue/diffusion-content-shift.

Not All Diffusion Model Activations Have Been Evaluated as Discriminative Features

Diffusion models are initially designed for image generation. Recent research shows that the internal signals within their backbones, named activations, can also serve as dense features for various discriminative tasks such as semantic segmentation. Given numerous activations, selecting a small yet effective subset poses a fundamental problem. To this end, the early study of this field performs a large-scale quantitative comparison of the discriminative ability of the activations. However, we find that many potential activations have not been evaluated, such as the queries and keys used to compute attention scores. Moreover, recent advancements in diffusion architectures bring many new activations, such as those within embedded ViT modules. Both combined, activation selection remains unresolved but overlooked. To tackle this issue, this paper takes a further step with a much broader range of activations evaluated. Considering the significant increase in activations, a full-scale quantitative comparison is no longer operational. Instead, we seek to understand the properties of these activations, such that the activations that are clearly inferior can be filtered out in advance via simple qualitative evaluation. After careful analysis, we discover three properties universal among diffusion models, enabling this study to go beyond specific models. On top of this, we present effective feature selection solutions for several popular diffusion models. Finally, the experiments across multiple discriminative tasks validate the superiority of our method over the SOTA competitors. Our code is available at https://github.com/Darkbblue/generic-diffusion-feature.

AUCSeg: AUC-oriented Pixel-level Long-tail Semantic Segmentation

The Area Under the ROC Curve (AUC) is a well-known metric for evaluating instance-level long-tail learning problems. In the past two decades, many AUC optimization methods have been proposed to improve model performance under long-tail distributions. In this paper, we explore AUC optimization methods in the context of pixel-level long-tail semantic segmentation, a much more complicated scenario. This task introduces two major challenges for AUC optimization techniques. On one hand, AUC optimization in a pixel-level task involves complex coupling across loss terms, with structured inner-image and pairwise inter-image dependencies, complicating theoretical analysis. On the other hand, we find that mini-batch estimation of AUC loss in this case requires a larger batch size, resulting in an unaffordable space complexity. To address these issues, we develop a pixel-level AUC loss function and conduct a dependency-graph-based theoretical analysis of the algorithm's generalization ability. Additionally, we design a Tail-Classes Memory Bank (T-Memory Bank) to manage the significant memory demand. Finally, comprehensive experiments across various benchmarks confirm the effectiveness of our proposed AUCSeg method. The code is available at https://github.com/boyuh/AUCSeg.

Improved Diversity-Promoting Collaborative Metric Learning for Recommendation

Collaborative Metric Learning (CML) has recently emerged as a popular method in recommendation systems (RS), closing the gap between metric learning and collaborative filtering. Following the convention of RS, existing practices exploit unique user representation in their model design. This paper focuses on a challenging scenario where a user has multiple categories of interests. Under this setting, the unique user representation might induce preference bias, especially when the item category distribution is imbalanced. To address this issue, we propose a novel method called \textit{Diversity-Promoting Collaborative Metric Learning} (DPCML), with the hope of considering the commonly ignored minority interest of the user. The key idea behind DPCML is to introduce a set of multiple representations for each user in the system where users' preference toward an item is aggregated by taking the minimum item-user distance among their embedding set. Specifically, we instantiate two effective assignment strategies to explore a proper quantity of vectors for each user. Meanwhile, a \textit{Diversity Control Regularization Scheme} (DCRS) is developed to accommodate the multi-vector representation strategy better. Theoretically, we show that DPCML could induce a smaller generalization error than traditional CML. Furthermore, we notice that CML-based approaches usually require \textit{negative sampling} to reduce the heavy computational burden caused by the pairwise objective therein. In this paper, we reveal the fundamental limitation of the widely adopted hard-aware sampling from the One-Way Partial AUC (OPAUC) perspective and then develop an effective sampling alternative for the CML-based paradigm. Finally, comprehensive experiments over a range of benchmark datasets speak to the efficacy of DPCML. Code are available at \url{https://github.com/statusrank/LibCML}.

Decorrelating Structure via Adapters Makes Ensemble Learning Practical for Semi-supervised Learning

In computer vision, traditional ensemble learning methods exhibit either a low training efficiency or the limited performance to enhance the reliability of deep neural networks. In this paper, we propose a lightweight, loss-function-free, and architecture-agnostic ensemble learning by the Decorrelating Structure via Adapters (DSA) for various visual tasks. Concretely, the proposed DSA leverages the structure-diverse adapters to decorrelate multiple prediction heads without any tailed regularization or loss. This allows DSA to be easily extensible to architecture-agnostic networks for a range of computer vision tasks. Importantly, the theoretically analysis shows that the proposed DSA has a lower bias and variance than that of the single head based method (which is adopted by most of the state of art approaches). Consequently, the DSA makes deep networks reliable and robust for the various real-world challenges, \textit{e.g.}, data corruption, and label noises. Extensive experiments combining the proposed method with FreeMatch achieved the accuracy improvements of 5.35% on CIFAR-10 dataset with 40 labeled data and 0.71% on CIFAR-100 dataset with 400 labeled data. Besides, combining the proposed method with DualPose achieved the improvements in the Percentage of Correct Keypoints (PCK) by 2.08% on the Sniffing dataset with 100 data (30 labeled data), 5.2% on the FLIC dataset with 100 data (including 50 labeled data), and 2.35% on the LSP dataset with 200 data (100 labeled data).