8822 papers
The paper proposes SpectralMamba-UNet, a frequency-disentangled framework that leverages discrete cosine transform to decouple and model low-frequency structural contexts and high-frequency boundary details via specialized state space mechanisms, achieving superior performance in medical image segmentation across diverse modalities.
The paper proposes FLIGHT, a novel real-time method that estimates camera heading from monocular video by generalizing the Hough transform on a unit sphere using a Fibonacci lattice to discretize directions, thereby achieving high accuracy and efficiency in noisy, outlier-rich environments while improving SLAM initialization.
This paper addresses the lack of standardized evaluation for adversarial transferability in image classification by reviewing existing methods, categorizing them into six groups, and proposing a comprehensive benchmark framework to ensure fair and unbiased comparisons while also discussing strategies and broader applications.
TriLite is a parameter-efficient, single-stage weakly supervised object localization framework that utilizes a frozen self-supervised ViT backbone and a novel TriHead module to achieve state-of-the-art performance with minimal trainable parameters by effectively disentangling foreground, background, and ambiguous regions.
The paper proposes CARE, a two-stage framework that addresses noisy-label person Re-Identification by first dismantling softmax overconfidence through probabilistic evidence calibration and then refining sample selection via a composite angular margin and certainty-oriented sphere weighting to effectively distinguish clean hard positives from mislabeled data.
This paper presents an unsupervised, online video stabilization framework that leverages classical priors and multithreaded buffering to overcome data and hardware limitations of deep learning methods, while introducing a new multimodal UAV dataset to validate its superior performance in challenging domains like nighttime remote sensing.
This study demonstrates that a computationally feasible approach combining coarse atmospheric data with sparse surface observations and high-resolution Earth observation can statistically recover physically coherent, meter-scale near-surface weather fields across the contiguous United States, significantly improving accuracy over existing reanalysis products.
This paper proposes BaryIR, a novel representation learning framework that achieves robust generalized all-in-one image restoration by decoupling degradation-agnostic invariant features in a Wasserstein barycenter space from degradation-specific residuals, thereby enabling effective adaptation to unseen degradations and real-world scenarios.
This paper presents Latent Gaussian Splatting (LaGS), a novel approach that unifies camera-based end-to-end tracking with mask-based multi-view panoptic occupancy prediction via a sparse 3D Gaussian latent representation to achieve state-of-the-art 4D spatiotemporal scene understanding on the Occ3D nuScenes and Waymo datasets.
This paper presents Phys-3D, a real-time physics-constrained tracking framework that integrates a transfer-learned YOLOv11m detector with a 3D motion model to achieve robust crowd counting on railway platforms despite challenges like camera motion, occlusions, and perspective distortions.
This paper introduces FairQuant, a framework for medical image classification that optimizes mixed-precision quantization to simultaneously maximize model efficiency under strict bit budgets and enhance algorithmic fairness across different demographic groups.
The paper proposes ColoDiff, a diffusion-based framework that generates high-quality, dynamic-consistent, and content-aware colonoscopy videos through novel inter-frame and intra-frame modules, effectively addressing data scarcity and enhancing downstream clinical tasks like disease diagnosis and lesion segmentation.
This paper presents a computational framework using the BrokenEyes system to simulate five common eye disorders and analyze their specific impacts on deep learning feature representations, revealing critical disruptions in models trained under degraded visual conditions.
This paper proposes Dual-Coupled PnP Diffusion with Spectral Homogenization, a novel framework that restores dual variables to eliminate steady-state bias in medical image reconstruction while transforming structured residuals into statistically compliant noise, thereby resolving the bias-hallucination trade-off and achieving state-of-the-art fidelity with accelerated convergence.
This paper proposes Multidimensional Task Learning (MTL), a unified tensor-based framework utilizing Generalized Einstein MLPs to overcome the limitations of matrix-based architectures by enabling the direct, lossless manipulation of tensor-valued parameters for a broader, more principled formulation of diverse computer vision tasks.
UniScale is a unified, feed-forward framework that enables robust, metric-scale 3D reconstruction for robotic perception by flexibly integrating geometric priors and jointly estimating camera parameters, depth, and point maps without requiring training from scratch.
This paper demonstrates that Large Multimodal Models (LMMs) can serve as effective general in-context classifiers, matching or surpassing contrastive VLMs in closed-world settings and establishing a robust training-free baseline called CIRCLE for open-world classification by iteratively refining pseudo-labels.
This paper argues that utilizing multiple camera views to generate more accurate 3D skeletons significantly enhances the performance of state-of-the-art action recognition models, suggesting that multi-view setups should become the standard for future research due to their favorable cost-benefit ratio.
The paper introduces GUIPruner, a training-free framework that employs Temporal-Adaptive Resolution and Stratified Structure-aware Pruning to eliminate spatiotemporal redundancy in high-resolution GUI agents, achieving significant efficiency gains and speedups while maintaining over 94% of original performance.
This paper proposes Risk-aware World Model Predictive Control (RaWMPC), a unified framework that enhances the generalization and safety of end-to-end autonomous driving by leveraging a risk-aware world model and self-evaluation distillation to make reliable decisions in unseen scenarios without relying on expert demonstrations.