4206 papers
EvoPrune is an early-stage visual token pruning method that performs layer-wise pruning guided by token similarity, diversity, and attention importance during visual encoding, achieving a 2 inference speedup with minimal performance degradation on high-resolution images and videos.
This paper proposes a wavelet-based polyp segmentation model that integrates grayscale and RGB representations through complementary frequency-consistent interaction to overcome low-contrast challenges and achieve superior boundary precision, as validated by extensive experiments on four benchmark datasets.
This paper proposes Embedded Runge-Kutta Guidance (ERK-Guid), a novel sampling method that leverages solver-induced local truncation errors as a guidance signal to detect stiffness and stabilize diffusion model generation, thereby outperforming state-of-the-art methods on benchmarks like ImageNet.
MPFlow is a zero-shot multi-modal MRI reconstruction framework that leverages a self-supervised pretraining strategy (PAMRI) to guide rectified flow sampling with auxiliary structural scans, thereby significantly reducing hallucinations and improving anatomical fidelity compared to single-modality baselines while requiring fewer sampling steps.
This paper identifies and quantifies "Order-to-Space Bias" (OTS), a systematic flaw in modern image generation models where the textual order of entities incorrectly dictates their spatial layout, and demonstrates that this data-driven issue can be effectively mitigated through targeted fine-tuning and early-stage interventions without compromising generation quality.
This paper introduces a novel dual-backbone architecture that fuses general visual features from a frozen DINOv3 model with task-specific features from a supervised Swin model to achieve state-of-the-art glass segmentation performance across multiple datasets while maintaining competitive inference speed.
To address the generalization challenges in No-Reference Point Cloud Quality Assessment caused by data scarcity, this paper proposes QD-PCQA, a novel unsupervised domain adaptation framework that transfers quality priors from images to point clouds through a Rank-weighted Conditional Alignment strategy and a Quality-guided Feature Augmentation module to enhance perceptual quality ranking and feature alignment.
The paper proposes PROSPECT, a unified streaming vision-language navigation agent that integrates CUT3R-based spatial encoding with SigLIP semantic features and employs latent predictive representation learning to achieve state-of-the-art performance and robustness in long-horizon navigation tasks.
DAGE introduces a dual-stream transformer architecture that efficiently estimates accurate, view-consistent geometry and camera poses from uncalibrated multi-view inputs by disentangling global coherence in a low-resolution stream from fine details in a high-resolution stream, achieving state-of-the-art performance while supporting high resolutions and long sequences.
This paper proposes WSI-INR, a novel patch-free framework utilizing Implicit Neural Representations and multi-resolution hash grid encoding to model whole-slide images as continuous functions, thereby overcoming the spatial fragmentation and resolution sensitivity of existing methods to achieve robust and accurate lesion segmentation across varying scales.
The paper introduces KFRA, a knowledge-augmented agent that emulates expert analysis through a three-stage closed reasoning loop to achieve superior open-set fine-grained visual understanding and interpretable, evidence-driven reasoning, validated by the newly constructed FGExpertBench.
DriveMVS is a novel multi-view stereo framework for autonomous driving that leverages sparse LiDAR observations as geometric prompts and employs a spatio-temporal decoder to achieve state-of-the-art metric accuracy, temporal consistency, and cross-domain generalization.
This paper proposes a novel framework for small object detection in complex backgrounds that integrates Residual Haar Wavelet Downsampling, Global Relation Modeling, Cross-Scale Hybrid Attention, and a Center-Assisted Loss to preserve fine-grained details, suppress noise, and enhance localization accuracy, achieving state-of-the-art performance on the RGBT-Tiny benchmark.
This paper introduces TAP, a training-free framework that accelerates diffusion model inference by adaptively selecting the most accurate predictor for each token at every step based on a low-cost probe, achieving significant speedups with minimal quality loss.
To address model collapse and knowledge loss in long-term test-time adaptation, this paper proposes an Adaptive and Selective Reset (ASR) framework that dynamically determines optimal reset timing and scope while employing an importance-aware regularizer to recover essential knowledge and an on-the-fly adjustment scheme to enhance adaptability.
The paper introduces STAR, a novel autoregressive pretraining method that utilizes image separators to quadruple the input sequence length of Vision Mamba, achieving a competitive 83.5% accuracy on ImageNet-1k by effectively leveraging long-range dependencies.
This paper proposes a lightweight underwater object detection framework based on YOLOv10 that integrates a Multi-Stage Adaptive Enhancement module, a Dual-Pooling Sequential Attention mechanism, and a Focal Generalized IoU loss to significantly improve accuracy and robustness on benchmark datasets while maintaining a compact model size suitable for resource-constrained environments.
This paper addresses the significant storage overhead of soft labels in dataset distillation by introducing a vector-quantized autoencoder (VQAE) that achieves 30–40x additional compression on benchmarks like ImageNet-1K while preserving over 90% of the original model performance.
This paper proposes Structure-aware Prompt Adaptation (SPA), a plug-and-play method that leverages the consistent local structures of semantically related concepts in the embedding space to effectively generalize from seen to unseen attributes and objects in Open-Vocabulary Compositional Zero-Shot Learning.
This paper identifies "Lazy Attention Localization" as a key bottleneck in multimodal cold-start training, where models fail to increase visual attention, and proposes the Attention-Guided Visual Anchoring and Reflection (AVAR) framework to effectively reshape attention distributions, achieving a 7.0% performance gain on multimodal reasoning benchmarks.