6474 papers
This paper introduces TraqPoint, a novel end-to-end reinforcement learning framework that reframes keypoint detection as a sequential decision-making problem to directly optimize trackability across image sequences, significantly outperforming existing state-of-the-art methods in 3D vision tasks.
This paper introduces Concept Distillation Sampling (CDS), a novel training-free framework that enables high-fidelity, multi-concept image editing by integrating a stable distillation backbone with dynamic weighting and spatially-aware LoRA priors to overcome the limitations of text-based optimization methods.
Uni-Animator is a novel Diffusion Transformer-based framework that unifies image and video sketch colorization by introducing instance patch embedding for precise color transfer, physical feature reinforcement for high-frequency detail preservation, and sketch-based dynamic RoPE encoding to ensure temporal coherence in large-motion scenes.
This paper introduces MedMAP, a modality-aware pre-training framework and the accompanying MedMoM-MRI3D dataset, which collectively enhance vision-language representation learning to significantly improve 3D MRI multi-organ abnormality detection across twelve modalities.
The paper proposes APPO, an attention-guided policy optimization algorithm that leverages token-level dense rewards to enhance fine-grained video perception through reasoning, demonstrating that improving perception yields greater performance gains than scaling reasoning capabilities alone.
This paper presents a novel approach to automating the curation and retrieval of 16th and 17th-century shipbuilding treatises by leveraging generative AI models like SAM2, Florence2, and ChatGPT, enhanced with specialized nautical ontologies, to overcome the challenges of segmenting and labeling scarce, domain-specific historical documents.
This paper presents Lilium, an automated evolutionary method that enhances the accuracy and robustness of craniofacial superimposition by modeling soft-tissue variability with a 3D cone-based representation optimized via Differential Evolution while enforcing multiple anatomical and photographic constraints.
GLIDE-Reg is a robust deformable registration method that jointly optimizes a registration field with a learnable dimensionality reduction module to fuse global semantic cues from foundation models with local handcrafted descriptors, achieving state-of-the-art performance in anatomical alignment and nodule tracking across diverse medical imaging cohorts.
This paper proposes IDER, a novel continual learning framework that leverages an idempotence distillation loss to seamlessly integrate uncertainty awareness with existing replay methods, thereby effectively reducing catastrophic forgetting while improving prediction reliability and accuracy.
The paper introduces BornoViT, a novel and highly efficient lightweight Vision Transformer model with only 0.65 million parameters that achieves 95.77% accuracy on the BanglaLekha dataset for classifying Bengali handwritten characters and digits, offering a resource-friendly alternative to computationally expensive state-of-the-art models.
ShiftLUT is an efficient image restoration framework that achieves a significantly larger receptive field and improved performance over state-of-the-art methods by integrating a Learnable Spatial Shift module, an asymmetric dual-branch architecture, and an Error-bounded Adaptive Sampling compression strategy, all while maintaining low storage and inference costs suitable for edge devices.
This paper introduces the Multimodal Weight Predictor (MWP) framework and the Waste-Weight-10K dataset to accurately estimate commercial and industrial waste weight by fusing RGB images with physics-informed metadata through a Vision Transformer and Stacked Mutual Attention, achieving high precision across a wide weight range while providing human-readable explanations via SHAP and large language models.
PreciseCache is a plug-and-play framework that significantly accelerates video generation by employing Low-Frequency Difference-based step-wise caching (LFCache) and block-wise caching (BlockCache) to precisely identify and skip truly redundant computations, thereby achieving substantial speedups without sacrificing output quality.
The paper introduces FlowReg, a flow-matching framework for unsupervised cardiac MR registration that utilizes warmup-reflow training and an Initial Guess strategy to achieve state-of-the-art accuracy with as few as two inference steps, eliminating the need for pre-trained models or segmentation labels.
HeroGS is a unified framework that enhances robust 3D Gaussian Splatting under sparse-view conditions by employing a hierarchical guidance strategy across image, feature, and parameter levels to regularize Gaussian distributions, refine high-frequency details, and ensure geometric consistency, thereby achieving superior reconstruction fidelity compared to state-of-the-art methods.
This paper introduces ET-Turb, a large-scale synthetic dataset and a novel exposure-time-dependent modulation transfer function (ET-MTF) framework that models atmospheric turbulence blur as a continuous function of exposure time, thereby enabling more realistic turbulence synthesis and significantly improving the generalization of vision models on real-world data compared to existing methods.
UETrack is a unified and efficient single object tracking framework that leverages a Token-Pooling-based Mixture-of-Experts mechanism and Target-aware Adaptive Distillation to achieve superior speed-accuracy trade-offs across multiple modalities and hardware platforms.
FACE introduces a novel face-level autoregressive framework that treats each triangle as a single token to drastically reduce sequence length and computational costs while achieving state-of-the-art reconstruction quality and enabling efficient high-fidelity 3D mesh generation.
This paper presents InterCoG, a novel text-vision interleaved chain-of-grounding reasoning framework that enhances fine-grained image editing in complex multi-entity scenes by explicitly deducing target locations through text-based spatial reasoning before performing visual grounding and outcome specification, supported by a new dataset and auxiliary training modules to ensure spatial precision.
This paper introduces BiCAM, a bidirectional class activation mapping method that captures both supportive and suppressive contributions in Vision Transformers to enhance explanation faithfulness and enable efficient adversarial detection without retraining.