The vanilla fusion methods still dominate a large percentage of mainstream audio-visual tasks. However, the effectiveness of vanilla fusion from a theoretical perspective is still worth discussing. Thus, this paper reconsiders the signal fused in the multimodal case from a bionics perspective and proposes a simple, plug-and-play, attention module for vanilla fusion based on fundamental signal theory and uncertainty theory. In addition, previous work on multimodal dynamic gradient modulation still relies on decoupling the modalities. So, a decoupling-free gradient modulation scheme has been designed in conjunction with the aforementioned attention module, which has various advantages over the decoupled one. Experiment results show that just a few lines of code can achieve up to 2.0% performance improvements to several multimodal classification methods. Finally, quantitative evaluation of other fusion tasks reveals the potential for additional application scenarios.
With the development of AI-Generated Content (AIGC), text-to-audio models are gaining widespread attention. However, it is challenging for these models to generate audio aligned with human preference due to the inherent information density of natural language and limited model understanding ability. To alleviate this issue, we formulate the BATON, a framework designed to enhance the alignment between generated audio and text prompt using human preference feedback. Our BATON comprises three key stages: Firstly, we curated a dataset containing both prompts and the corresponding generated audio, which was then annotated based on human feedback. Secondly, we introduced a reward model using the constructed dataset, which can mimic human preference by assigning rewards to input text-audio pairs. Finally, we employed the reward model to fine-tune an off-the-shelf text-to-audio model. The experiment results demonstrate that our BATON can significantly improve the generation quality of the original text-to-audio models, concerning audio integrity, temporal relationship, and alignment with human preference.
Existing learned video compression models employ flow net or deformable convolutional networks (DCN) to estimate motion information. However, the limited receptive fields of flow net and DCN inherently direct their attentiveness towards the local contexts. Global contexts, such as large-scale motions and global correlations among frames are ignored, presenting a significant bottleneck for capturing accurate motions. To address this issue, we propose a joint local and global motion compensation module (LGMC) for leaned video coding. More specifically, we adopt flow net for local motion compensation. To capture global context, we employ the cross attention in feature domain for motion compensation. In addition, to avoid the quadratic complexity of vanilla cross attention, we divide the softmax operations in attention into two independent softmax operations, leading to linear complexity. To validate the effectiveness of our proposed LGMC, we integrate it with DCVC-TCM and obtain learned video compression with joint local and global motion compensation (LVC-LGMC). Extensive experiments demonstrate that our LVC-LGMC has significant rate-distortion performance improvements over baseline DCVC-TCM.
Stochastic optimization methods encounter new challenges in the realm of streaming, characterized by a continuous flow of large, high-dimensional data. While first-order methods, like stochastic gradient descent, are the natural choice, they often struggle with ill-conditioned problems. In contrast, second-order methods, such as Newton's methods, offer a potential solution, but their computational demands render them impractical. This paper introduces adaptive stochastic optimization methods that bridge the gap between addressing ill-conditioned problems while functioning in a streaming context. Notably, we present an adaptive inversion-free Newton's method with a computational complexity matching that of first-order methods, $\mathcal{O}(dN)$, where $d$ represents the number of dimensions/features, and $N$ the number of data. Theoretical analysis confirms their asymptotic efficiency, and empirical evidence demonstrates their effectiveness, especially in scenarios involving complex covariance structures and challenging initializations. In particular, our adaptive Newton's methods outperform existing methods, while maintaining favorable computational efficiency.
In the realm of robot-assisted minimally invasive surgery, dynamic scene reconstruction can significantly enhance downstream tasks and improve surgical outcomes. Neural Radiance Fields (NeRF)-based methods have recently risen to prominence for their exceptional ability to reconstruct scenes. Nonetheless, these methods are hampered by slow inference, prolonged training, and substantial computational demands. Additionally, some rely on stereo depth estimation, which is often infeasible due to the high costs and logistical challenges associated with stereo cameras. Moreover, the monocular reconstruction quality for deformable scenes is currently inadequate. To overcome these obstacles, we present Endo-4DGS, an innovative, real-time endoscopic dynamic reconstruction approach that utilizes 4D Gaussian Splatting (GS) and requires no ground truth depth data. This method extends 3D GS by incorporating a temporal component and leverages a lightweight MLP to capture temporal Gaussian deformations. This effectively facilitates the reconstruction of dynamic surgical scenes with variable conditions. We also integrate Depth-Anything to generate pseudo-depth maps from monocular views, enhancing the depth-guided reconstruction process. Our approach has been validated on two surgical datasets, where it can effectively render in real-time, compute efficiently, and reconstruct with remarkable accuracy. These results underline the vast potential of Endo-4DGS to improve surgical assistance.
We propose EnCLAP, a novel framework for automated audio captioning. EnCLAP employs two acoustic representation models, EnCodec and CLAP, along with a pretrained language model, BART. We also introduce a new training objective called masked codec modeling that improves acoustic awareness of the pretrained language model. Experimental results on AudioCaps and Clotho demonstrate that our model surpasses the performance of baseline models. Source code will be available at //github.com/jaeyeonkim99/EnCLAP . An online demo is available at //huggingface.co/spaces/enclap-team/enclap .
Over the past few years, the rapid development of deep learning technologies for computer vision has greatly promoted the performance of medical image segmentation (MedISeg). However, the recent MedISeg publications usually focus on presentations of the major contributions (e.g., network architectures, training strategies, and loss functions) while unwittingly ignoring some marginal implementation details (also known as "tricks"), leading to a potential problem of the unfair experimental result comparisons. In this paper, we collect a series of MedISeg tricks for different model implementation phases (i.e., pre-training model, data pre-processing, data augmentation, model implementation, model inference, and result post-processing), and experimentally explore the effectiveness of these tricks on the consistent baseline models. Compared to paper-driven surveys that only blandly focus on the advantages and limitation analyses of segmentation models, our work provides a large number of solid experiments and is more technically operable. With the extensive experimental results on both the representative 2D and 3D medical image datasets, we explicitly clarify the effect of these tricks. Moreover, based on the surveyed tricks, we also open-sourced a strong MedISeg repository, where each of its components has the advantage of plug-and-play. We believe that this milestone work not only completes a comprehensive and complementary survey of the state-of-the-art MedISeg approaches, but also offers a practical guide for addressing the future medical image processing challenges including but not limited to small dataset learning, class imbalance learning, multi-modality learning, and domain adaptation. The code has been released at: //github.com/hust-linyi/MedISeg
Video captioning is a challenging task that requires a deep understanding of visual scenes. State-of-the-art methods generate captions using either scene-level or object-level information but without explicitly modeling object interactions. Thus, they often fail to make visually grounded predictions, and are sensitive to spurious correlations. In this paper, we propose a novel spatio-temporal graph model for video captioning that exploits object interactions in space and time. Our model builds interpretable links and is able to provide explicit visual grounding. To avoid unstable performance caused by the variable number of objects, we further propose an object-aware knowledge distillation mechanism, in which local object information is used to regularize global scene features. We demonstrate the efficacy of our approach through extensive experiments on two benchmarks, showing our approach yields competitive performance with interpretable predictions.
Deep neural networks (DNNs) are successful in many computer vision tasks. However, the most accurate DNNs require millions of parameters and operations, making them energy, computation and memory intensive. This impedes the deployment of large DNNs in low-power devices with limited compute resources. Recent research improves DNN models by reducing the memory requirement, energy consumption, and number of operations without significantly decreasing the accuracy. This paper surveys the progress of low-power deep learning and computer vision, specifically in regards to inference, and discusses the methods for compacting and accelerating DNN models. The techniques can be divided into four major categories: (1) parameter quantization and pruning, (2) compressed convolutional filters and matrix factorization, (3) network architecture search, and (4) knowledge distillation. We analyze the accuracy, advantages, disadvantages, and potential solutions to the problems with the techniques in each category. We also discuss new evaluation metrics as a guideline for future research.
Many tasks in natural language processing can be viewed as multi-label classification problems. However, most of the existing models are trained with the standard cross-entropy loss function and use a fixed prediction policy (e.g., a threshold of 0.5) for all the labels, which completely ignores the complexity and dependencies among different labels. In this paper, we propose a meta-learning method to capture these complex label dependencies. More specifically, our method utilizes a meta-learner to jointly learn the training policies and prediction policies for different labels. The training policies are then used to train the classifier with the cross-entropy loss function, and the prediction policies are further implemented for prediction. Experimental results on fine-grained entity typing and text classification demonstrate that our proposed method can obtain more accurate multi-label classification results.
We construct targeted audio adversarial examples on automatic speech recognition. Given any audio waveform, we can produce another that is over 99.9% similar, but transcribes as any phrase we choose (at a rate of up to 50 characters per second). We apply our iterative optimization-based attack to Mozilla's implementation DeepSpeech end-to-end, and show it has a 100% success rate. The feasibility of this attack introduce a new domain to study adversarial examples.