论文检索

Kui Jiang, Jing Cao, Zhaocheng Yu, Junjun Jiang, Jingchun Zhou

Monocular depth estimation is critical for applications such as autonomous driving and scene reconstruction. While existing methods perform well under normal scenarios, their performance declines in adverse weather, due to challenging domain shifts and difficulties in extracting scene information. To address this issue, we present a robust monocular depth estimation method called ACDepth from the perspective of high-quality training data generation and domain adaptation. Specifically, we introduce a one-step diffusion model for generating samples that simulate adverse weather conditions, constructing a multi-tuple degradation dataset during training. To ensure the quality of the generated degradation samples, we employ LoRA adapters to fine-turn the generation weights of diffusion model. Additionally, we integrate circular consistency loss and adversarial training to guarantee the fidelity and naturalness of the scene contents. Furthermore, we elaborate on a multi-granularity knowledge distillation strategy (MKD) that encourages the student network to absorb knowledge from both the teacher model and pretrained Depth Anything V2. This strategy guides the student model in learning degradation-agnostic scene information from various degradation inputs. In particular, we introduce an ordinal guidance distillation mechanism (OGD) that encourages the network to focus on uncertain regions through differential ranking, leading to a more precise depth estimation. Experimental results demonstrate that our ACDepth surpasses md4all-DD by 2.50% for night scene and 2.61% for rainy scene on the nuScenes dataset in terms of the absRel metric.

Jianchao Jiang, Haofeng Zhang

Few-Shot Medical Image Segmentation (FSMIS) has been widely used to train a model that can perform segmentation from only a few annotated images. However, most existing prototype-based FSMIS methods generate multiple prototypes from the support image solely by random sampling or local averaging, which can cause particularly severe boundary blurring due to the tendency for normal features accounting for the majority of features of a specific category. Consequently, we propose to focus more attention to those weaker features that are crucial for clear segmentation boundary. Specifically, we design a Support Self-Prediction (SSP) module to identify such weak features by comparing true support mask with one predicted by global support prototype. Then, a Hard Prototypes Generation (HPG) module is employed to generate multiple hard prototypes based on these weak features. Subsequently, a Multiple Similarity Maps Fusion (MSMF) module is devised to generate final segmenting mask in a dual-path fashion to mitigate the imbalance between foreground and background in medical images. Furthermore, we introduce a boundary loss to further constraint the edge of segmentation. Extensive experiments on three publicly available medical image datasets demonstrate that our method achieves state-of-the-art performance. Code is available at https://github.com/jcjiang99/CoW.

Ran Jia, Junpeng Xue, Shuai Ma, Wenbo Lu, Kelei Wang

Point cloud completion is a crucial task in 3D computer vision. Multi-modal completion approaches have gained attention among the popular two-stage point cloud completion methods. However, there is a notable lack of research focused on accurately aligning data from different modalities within these methods. Additionally, in other point cloud-based tasks, edge point information often provides unexpected positive contributions. In this paper, we propose a novel point cloud completion method that leverages edge point information for the first time in the completion task, which also addresses the precise alignment of multi-modal data. In particular, we implement a two-step local-to-global module to achieve better alignment of multi-modal data during the preliminary point cloud generation process. Besides, we introduce a new spatial representation structure capable of extracting a fixed number of edge points. Moreover, with the assistance of edge information, we further design an inverse edge-aware upsampler to refine the point cloud. We evaluate our method on three typical datasets, and the results demonstrate that our IE-PMMA outperforms the existing state-of-the-art methods quantitatively and visually.

Yuzhu Ji, Mingshan Sun, Jianyang Shi, Xiaoke Jiang, Yiqun Zhang, Haijun Zhang

Category-level object pose estimation is a longstanding and fundamental task crucial for augmented reality and robotic manipulation applications. Existing RGB-based approaches struggle with multi-stage settings and heavily rely on off-the-shelf techniques, such as object detectors, depth estimators, non-differentiable NOCS shape alignment, etc. Extra dependencies lead to the accumulation of errors and complicate the whole pipeline, limiting the deployment of these approaches in practical applications. This paper streamlined an end-to-end framework unifying the single-frame and video-based category-level pose estimation. Specifically, instead of explicitly introducing extra dependencies, the DINOv2 encoder and depth decoder, as robust semantic and geometric prior extractors, are leveraged to produce intra-frame hierarchical semantic and geometric features. A spatial-temporal sparse query network is developed to model the implicit correspondence and inter-frame correlations between a set of implicit 3D query anchors and intra-frame features. Finally, a pose prediction head is employed using the bipartite matching algorithm. Experimental results demonstrate that our model achieves state-of-the-art performance compared with RGB-based categorical pose estimation methods on the REAL275 and CAMERA25 datasets. Our code is available at https://andrewchiyz.github.io/vision.3dv.seqpose/.

Xueyao Ji, Gang Wang, Yizheng Wang

Optical flow is pivotal in video-based tasks, yet existing methods mostly focus on medium-/large-size objects, while underperforming when characterizing the motion of tiny objects. To bridge this gap, we introduce the On-off Time-delay with Hassenstein-Reichardt correlator (OTHR), a computationally efficient scheme inspired by the primate visual cortex's direction selectivity mechanism. OTHR kernels, applied across multiple frames, discern bright/dark luminance changes along a specific direction over a time delay, effectively estimating motion of tiny objects amidst noise and static backgrounds. Notably, OTHR integrates seamlessly with leading deep learning flow estimation models such as RAFT and FlowFormer. We also propose refined evaluation metrics for tiny objects and contribute a new dataset featuring such objects to aid algorithm development. Our experiments confirm OTHR's superiority over competing methods, particularly in enhancing state-of-the-art models' performance on tiny object motion estimation at minimal cost. Specifically, for objects less than 100 pixels, OTHR reduces RAFT and FlowFormer's errors by 22.03% and 83.50%, respectively. The codes will be accessible at https://github.com/JaneEliot/OTHR.

Andela Ilic, Jiaxi Jiang, Paul Streli, Xintong Liu, Christian Holz

Motion capture using sparse inertial sensors has shown great promise due to its portability and lack of occlusion issues compared to camera-based tracking. Existing approaches typically assume that IMU sensors are tightly attached to the human body. However, this assumption often does not hold in real-world scenarios. In this paper, we present a new task of full-body human pose estimation using sparse, loosely attached IMU sensors. To solve this task, we simulate IMU recordings from an existing garment-aware human motion dataset. We developed transformer-based diffusion models to synthesize loose IMU data and estimate human poses based on this challenging loose IMU data. In addition, we show that incorporating garment-related parameters while training the model on simulated loose data effectively maintains expressiveness and enhances the ability to capture variations introduced by looser or tighter garments. Experiments show that our proposed diffusion methods trained on simulated and synthetic data outperformed the state-of-the-art methods quantitatively and qualitatively, opening up a promising direction for future research.

Xingru Huang, Jian Huang, Yihao Guo, Tianyun Zhang, Zhao Huang, Yaqi Wang, Ruipu Tang, Guangliang Cheng, Shaowei Jiang, Zhiwen Zheng 等

Information retrieved from three dimensions is treated uniformly in CNN-based volumetric segmentation methods. However, such neglect of axial disparities fails to capture true spatio-temporal variations. This paper introduces the volumetric axial disentanglement to address the disparities in spatial information along different axial dimensions. Building on this concept, we propose the Post-Axial Refiner (PaR) module to refine segmentation masks by implementing axial disentanglement on the specific axis of the volumetric medical sequences. As a plug-and-play enhancement to existing volumetric segmentation architecture, PaR further utilizes specialized attention approaches to learn disentangled post-decoding features, enhancing spatial representation and structural detail. Validation on various datasets demonstrates PaR's consistent elevation of segmentation precision and boundary clarity across 11 baselines and different imaging modalities, achieving state-of-the-art performance on multiple datasets. Experimental tests demonstrate the ability of volumetric axial disentanglement to refine the segmentation of volumetric medical images. Code is released at https://github.com/IMOP-lab/PaR-Pytorch.

Jinglin Huang, Maoqiang Wu, Peichun Li, Wen Wu, Rong Yu

Cross-view geo-localization is a crucial task with diverse applications, yet it remains challenging due to the significant variations in viewpoints and visual appearances between images from different perspectives. While recent advancements have been made, existing methods often suffer from high model complexity, excessive resource consumption, and the impact of sample learning difficulty on optimization. To overcome these limitations, we optimize the Vision Mamba (Vim) model, built on a State Space Model (SSM) architecture, by replacing the traditional classification head with Channel Group Pooling (CGP) for efficient feature integration. This optimization reduces model parameters by 1.5% and computational complexity by 0.4%. Additionally, we propose a novel Dynamic Weighted Batch-tuple Loss (DWBL) to dynamically adjust the weighting of negative samples, improving model performance. By combining CGP and DWBL, we develop an efficient end-to-end network, VimGeo, which achieves state-of-the-art performance with enhanced computational efficiency. Specifically, VimGeo achieves a Recall@1 of 81.67% on the CVACT_test dataset, outperforming prior approaches. Extensive experiments on CVUSA, CVACT, and VIGOR datasets validate VimGeo's effectiveness and competitiveness in cross-view geo-localization tasks, achieving the leading results among sequence modeling-based methods. The implementation is available at: https://github.com/VimGeoTeam/VimGeo.

He Huang, Qi Yang, Mufan Liu, Yiling Xu, Zhu Li

Existing 4D Gaussian Splatting methods rely on per-Gaussian deformation from a canonical space to target frames, which overlooks redundancy among adjacent Gaussian primitives and result in suboptimal performance. To address this limitation, we propose Anchor-Driven Deformable and Compressed Gaussian Splatting (ADC-GS), a compact and efficient representation for dynamic scene reconstruction. Specifically, ADC-GS organizes Gaussian primitives into an anchor-based structure within the canonical space, enhanced by a temporal significance-based anchor refinement strategy. To reduce deformation redundancy, ADC-GS introduces a hierarchical coarse-to-fine pipeline that captures motions at varying granularities. Moreover, a rate-distortion optimization is adopted to achieve an optimal balance between bitrate consumption and representation fidelity. Experimental results demonstrate that ADC-GS outperforms the per-Gaussian deformation approaches in rendering speed by 300%-800% while achieving state-of-the-art storage efficiency without compromising rendering quality. The code is released at https://github.com/H-Huang774/ADC-GS.git.

Feizhen Huang, Yu Wu, Yutian Lin, Bo Du

Video-to-Audio (V2A) Generation achieves significant progress and plays a crucial role in film and video post-production. However, current methods overlook the cinematic language, a critical component of artistic expression in filmmaking. As a result, their performance deteriorates in scenarios where Foley targets are only partially visible. To address this challenge, we propose a simple self-distillation approach to extend V2A models to cinematic language scenarios. By simulating the cinematic language variations, the student model learns to align the video features of training pairs with the same audio-visual correspondences, enabling it to effectively capture the associations between sounds and partial visual information. Our method not only achieves impressive improvements under partial visibility across all evaluation metrics, but also enhances performance on the large-scale V2A dataset, VGGSound.

Chaowei Huang, Yaru Su, Huangbiao Xu, Xiao Ke

Multi-modality image fusion (MMIF) integrates features from distinct modalities to enhance visual quality and improve downstream task performance. However, existing methods often overlook the sparsity variations and dynamic correlations between infrared and visible images, potentially limiting the utilization of both modalities. To address these challenges, we propose the Progressive Modality-Adaptive Interactive Network (PoMAI), a novel framework that not only dynamically adapts to the sparsity and structural disparities of each modality but also enhances inter-modal correlations, thereby optimizing fusion quality. The training process consists of two stages: in the first stage, the Neighbor-Group Matching Model (NGMM) models the high sparsity of infrared features, while the Context-Aware Modeling Network (CAMN) captures rich structural details in visible features, jointly refining modality-specific characteristics for fusion. In the second stage, the Modality-Interactive Compensation Module (MICM) refines inter-modal correlations via dynamic compensation mechanism, while freezing the first-stage modules to focus MICM solely on the compensation task. Extensive experiments on benchmark datasets demonstrate that PoMAI surpasses state-of-the-art methods in fusion quality and excels in downstream tasks.

Chao Huang, Qianyi Li, Jie Wen, Bob Zhang

Pixel-level anomaly detection is indispensable in industrial defect detection and medical diagnosis. Recently, Segment Anything Model (SAM) has achieved promising results in many vision tasks. However, direct application of the SAM to pixel-level anomaly detection tasks results in unsatisfactory performance, meanwhile SAM needs the manual prompt. Although some automatically prompt-based SAM has been proposed, these automated prompting approaches merely utilize partial image features as prompts and fail to incorporate crucial features such as multi-scale image features to generate more suitable prompts. In this paper, we propose a novel Omni Dimensional State Space Model-driven SAM (ODS-SAM) for pixel-level anomaly detection. Specifically, the proposed method adopts the SAM architecture, ensuring easy implementation and avoiding the need for fine-tuning. A State-Space Model-based residual Omni Dimensional module is designed to automatically generate suitable prompts. This module can effectively leverage multi-scale and global information, facilitating an iterative search for optimal prompts in the prompt space. The identified optimal prompts are then fed into SAM as high-dimensional tensors. Experimental results demonstrate that the proposed ODS-SAM outperforms state-of-the-art models on both industrial and medical image datasets.

Binxiao Huang, Zhihao Li, Shiyong Liu, Xiao Tang, Jiajun Tang, Jiaqi Lin, Yuxin Cheng, Zhenyu Chen, Xiaofei Wu, Ngai Wong

3D Gaussian splatting (3DGS) has demonstrated exceptional performance in image-based 3D reconstruction and real-time rendering. However, regions with complex textures require numerous Gaussians to capture significant color variations accurately, leading to inefficiencies in rendering speed. To address this challenge, we introduce a hybrid representation for indoor scenes that combines 3DGS with textured meshes. Our approach uses textured meshes to handle texture-rich flat areas, while retaining Gaussians to model intricate geometries. The proposed method begins by pruning and refining the extracted mesh to eliminate geometrically complex regions. We then employ a joint optimization for 3DGS and mesh, incorporating a warm-up strategy and transmittance-aware supervision to balance their contributions seamlessly.Extensive experiments demonstrate that the hybrid representation maintains comparable rendering quality and achieves superior frames per second FPS with fewer Gaussian primitives.

Binxiao Huang, Ngai Wong

Poisoning-based backdoor attacks expose vulnerabilities during the data preparation phase of deep neural network (DNN) training. The DNNs trained on the poisoned dataset will be embedded with a backdoor, making them behave well on clean data while outputting malicious predictions whenever a trigger is applied. To exploit the abundant information contained in the input-to-label mapping, our scheme utilizes the network trained from the clean dataset as a trigger generator to produce poisons that significantly raise the success rate of backdoor attacks versus conventional approaches. Specifically, we introduce a new categorization of triggers inspired by adversarial techniques and propose a multi-label and multi-payload Poisoning-based backdoor attack with Positive Triggers (PPT), which strategically manipulates inputs to align them closer to the target label in the feature space of benign classifiers. Once the classifier is trained on the poisoned dataset, we can generate an input-label-aware trigger to make the infected classifier predict any given input to any target label with a high possibility. Through extensive experiments under both dirty-label and clean-label settings, we demonstrate empirically that the proposed attack achieves a high attack success rate without sacrificing accuracy across various datasets, including SVHN, CIFAR10, GTSRB, and Tiny ImageNet. Additionally, the PPT attack can elude a variety of classical backdoor defenses, proving its effectiveness.

Qingguo Hu, Ante Wang, Jia Song, Delai Qiu, Qingsong Liu, Jinsong Su

Large Vision-Language Models (LVLMs) have experienced significant advancements in recent years. However, their performance still falls short in tasks requiring deep visual perception, such as identifying subtle differences between images. A potential cause is the scarcity of visual knowledge in popular instruction-tuning corpora, resulting in inadequate visual perception and reasoning capabilities. To address this challenge, we introduce a self-improvement framework grounded in a novel visual knowledge-intensive task, Causality-driven Visual object Completion (CVC). This task requires LVLMs to infer the masked object in an image based on its causal relationships with the other visible information. We first obtain rich examples cheaply through our automated instance construction pipeline, without relying on sophisticated LVLMs (e.g., GPT-4V) or human assistance. Then, LVLMs effectively self-improve through trial and error learning using these created instances. Our experiments demonstrate substantial gains across four challenging specialized tasks and four widely-used comprehensive benchmarks. Especially on specialized tasks, our method achieves an average improvement of 5.4% and 4.0% compared to the corresponding baselines when utilizing LLaVA-1.5-7B and LLaVA-1.5-13B, respectively. Code and the supplementary file are available at https://github.com/XMUDeepLIT/CVC.

Jing Hu, Chengming Feng, Shu Hu, Ming-Ching Chang, Xin Li, Xi Wu, Xin Wang

Arbitrary style transfer aims to apply the style of any given artistic image to another content image. Still, existing deep learning-based methods often require significant computational costs to generate diverse stylized results. Motivated by this, we propose a novel reinforcement learning-based framework for arbitrary style transfer RLMiniStyler. This framework leverages a unified reinforcement learning policy to iteratively guide the style transfer process by exploring and exploiting stylization feedback, generating smooth sequences of stylized results while achieving model lightweight. Furthermore, we introduce an uncertainty-aware multi-task learning strategy that automatically adjusts loss weights to adapt to the content and style balance requirements at different training stages, thereby accelerating model convergence. Through a series of experiments across image various resolutions, we have validated the advantages of RLMiniStyler over other state-of-the-art methods in generating high-quality, diverse artistic image sequences at a lower cost. Codes are available at https://github.com/fengxiaoming520/RLMiniStyler.

Jian Hu, Zixu Cheng, Shaogang Gong

Task-generic promptable image segmentation aims to achieve segmentation of diverse samples under a single task description by utilizing only one task-generic prompt. Current methods leverage the generalization capabilities of Vision-Language Models (VLMs) to infer instance-specific prompts from these task-generic prompts in order to guide the segmentation process. However, when VLMs struggle to generalise to some image instances, predicting instance-specific prompts becomes poor. To solve this problem, we introduce Instance-specific Negative Mining for Task-Generic Promptable Segmentation (INT). The key idea of INT is to adaptively reduce the influence of irrelevant (negative) prior knowledge whilst to increase the use the most plausible prior knowledge, selected by negative mining with higher contrast, in order to optimise instance-specific prompts generation. Specifically, INT consists of two components: (1) instance-specific prompt generation, which progressively fliters out incorrect information in prompt generation; (2) semantic mask generation, which ensures each image instance segmentation matches correctly the semantics of the instance-specific prompts. INT is validated on six datasets, including camouflaged objects and medical images, demonstrating its effectiveness, robustness and scalability.

Xin He, Longhui Wei, Lingxi Xie, Qi Tian

Multimodal Large Language Models (MLLMs) are experiencing rapid growth, yielding a plethora of novel works recently. The prevailing trend involves adopting data-driven methodologies, wherein diverse instruction-following datasets were collected. However, these approaches always face the challenge of limited visual perception capabilities, as they solely utilizing CLIP-like encoders to extract visual information from inputs. Though these encoders are pre-trained on billions of image-text pairs, they still grapple with the information loss dilemma, given that textual captions only partially capture the contents depicted in images. To address this limitation, this paper proposes to improve the visual perception ability of MLLMs through a mixture-of-experts knowledge enhancement mechanism. Specifically, this work introduces a novel method that incorporates multi-task encoders and existing visual tools into the MLLMs training and inference pipeline, aiming to provide a more comprehensive summarization of visual inputs. Extensive experiments have evaluated its effectiveness of advancing MLLMs, showcasing improved visual perception capability achieved through the integration of visual experts.

Gang He, Kepeng Xu, Li Xu, Wenxin Yu, Xianyun Wu

The rise of HDR-WCG display devices has highlighted the need to convert SDRTV to HDRTV, as most video sources are still in SDR. Existing methods primarily focus on designing neural networks to learn a single-style mapping from SDRTV to HDRTV. However, the limited information in SDRTV and the diversity of styles in real-world conversions render this process an ill-posed problem, thereby constraining the performance and generalization of these methods. Inspired by generative approaches, we propose a novel method for SDRTV to HDRTV conversion guided by real HDRTV priors. Despite the limited information in SDRTV, introducing real HDRTV as reference priors significantly constrains the solution space of the originally high-dimensional ill-posed problem. This shift transforms the task from solving an unreferenced prediction problem to making a referenced selection, thereby markedly enhancing the accuracy and reliability of the conversion process. Specifically, our approach comprises two stages: the first stage employs a Vector Quantized Generative Adversarial Network to capture HDRTV priors, while the second stage matches these priors to the input SDRTV content to recover realistic HDRTV outputs. We evaluate our method on public datasets, demonstrating its effectiveness with significant improvements in both objective and subjective metrics across real and synthetic datasets.

Tianxiang Hao, Hui Chen, Guiguang Ding

In order to boost the performance of a convolutional neural network (CNN), several approaches have shown the benefit of enhancing the spatial encoding of feature maps. However, few works paid attention to the positional properties of convolutional kernels. In this paper, we demonstrate that different kernel positions are of different importance, which depends on the task, dataset and architecture, and adaptively emphasizing the informative parts in convolutional kernels can lead to considerable improvement. Therefore, we propose a novel structural re-parameterization Position Boosting Convolution (PBConv) to exploit and enhance the position information in the convolutional kernel. PBConv consists of several concurrent small convolutional kernels, which can be equivalently converted to the original kernel and bring no extra inference cost. Different from existing structural re-parameterization methods, PBconv searches for the optimal re-parameterized structure by a fast heuristic algorithm based on the dispersion of kernel weights. Such heuristic search is efficient yet effective, well adapting the varying kernel weight distribution. As a result, PBConv can significantly improve the representational power of a model, especially its ability to extract fine-grained low-level features. Importantly, PBConv is orthogonal to procedural re-parameterization methods and can further boost performance based on them.