LePaX enables high-resolution chest X-ray report generation by learning to allocate resolution to diagnostically relevant regions and fusing high-res patches back into global features without increasing token count.
Title resolution pending
44 Pith papers cite this work. Polarity classification is still indexing.
44
Pith papers citing it
citation-role summary
method 3
baseline 1
citation-polarity summary
representative citing papers
IS-SNN removes activation normalization from deep SNNs via topology-aware weight standardization folded into static weights, matching dynamic BN accuracy on ImageNet (68.05%) while cutting FPGA neuron LUT usage by 96.4%.
A^4D detects adversarial attacks in an attack- and classifier-agnostic way by measuring non-arbitrary shifts in CLIP embedding space from prompt-based similarity scores.
Introduces synthetic ground-truth dataset for CAM evaluation, proposes ARCC composite metric, and RefineCAM method that aggregates layers for higher-resolution maps outperforming baselines.
PromptDx adds a differentiable adapter to align multimodal data with a pre-trained TabPFN-style ICL engine, achieving strong Alzheimer's diagnosis performance with only 1% context samples.
MoCA3D formulates monocular 3D box prediction as dense pixel-space tasks using corner heatmaps and depth maps, with a new PAG metric for image-plane evaluation.
Hybrid T2I generation with teacher-student pseudo-labeling plus VRAIN context-aware I2I rare-class editing improves LVIS instance segmentation AP, especially on rare categories.
TPB is an AdaBoost-style ensemble method for text prompts in VLMs that improves few-shot accuracy by targeting hard examples and maintains gains across model transfers.
Echo4DIR reconstructs continuous 4D cardiac geometry from sparse 2D echocardiography videos using implicit representations, epipolar feature fusion, self-supervised domain adaptation, and radial SDF alignment to achieve up to 98.35% Dice overlap.
Mirage auditing framework reveals that VFL unlearning methods passing output-level certification retain substantial class structure in representations, with no method achieving high utility plus both output and representation forgetting, plus class-sample asymmetry in residual traces.
Pretrained autoencoders in medical latent diffusion encode discriminative features well for reconstruction but structure their latent spaces in ways that hinder classifier learning, a gap that persists across architectures and is not closed by domain fine-tuning.
A verifiable CBM framework grounds concepts in localized image patches, achieving comparable accuracy to standard CBMs on medical benchmarks while enabling direct inspection of concept correctness.
HashSCD is a patch-wise hashing method for unsupervised scene change detection and localization that operates directly in Hamming space with competitive performance and lower computational cost.
Counterfactual stress testing with causal generative models offers a more accurate proxy than simple perturbations for predicting medical image model performance under distribution shifts.
Hyp2Former learns hierarchical semantic similarities in hyperbolic space among known categories so that unknown objects remain close to higher-level concepts and can be detected reliably.
An asymmetric multi-level distillation framework lets a student ViT approximate clean-image representations from distorted inputs alone, outperforming prior methods on classification under distortions.
Logic Gate Networks produce compact Boolean-circuit descriptors for video copy detection that match or exceed prior accuracy at over 11k inferences per second and orders-of-magnitude smaller size.
LLMasTool improves neural architecture search by evolving code-mined hierarchical trees with diversity-guided Bayesian planning and targeted LLM assistance, reporting gains of 0.69, 1.83, and 2.68 points on CIFAR-10, CIFAR-100, and ImageNet16-120.
ESCAPE combines spatio-temporal fusion mapping for depth-free 3D memory with a memory-driven grounding module and adaptive execution policy to reach 65.09% success on ALFRED test-seen long-horizon mobile manipulation tasks.
A new high-resolution landscape concept dataset enables the first Robust TCAV explanations of deep SDMs, largely validating ecological expectations for Plecoptera and Trichoptera while flagging novel associations.
GroundingAnomaly uses a Spatial Conditioning Module and Gated Self-Attention in a frozen diffusion U-Net to synthesize spatially accurate few-shot anomalies, reaching SOTA on MVTec AD and VisA for detection, segmentation, and instance detection.
PRISM-CTG is the first large-scale foundation model for cardiotocography that uses multi-view self-supervised learning on unlabeled data to learn transferable representations, outperforming baselines on seven downstream tasks with external validation.
Hierarchical clinical-reasoning SFT plus counterfactual GRPO yields SoTA diagnostic accuracy for multimodal LLMs on gastrointestinal endoscopy benchmarks.
ShapeFuse uses bidirectional cross-modal temporal attention and adaptive gating to fuse deformable shape and texture features for cardiac video classification, outperforming existing fusion strategies on a cine CMR dataset.
citing papers explorer
-
Seeing What Matters: Lesion-Aware High-Resolution Patch Discovery and Fusion for Chest X-ray Report Generation
LePaX enables high-resolution chest X-ray report generation by learning to allocate resolution to diagnostically relevant regions and fusing high-res patches back into global features without increasing token count.
-
Intrinsically Stable Spiking Neural Networks: Overcoming the Performance Barrier in the Absence of Batch Normalization
IS-SNN removes activation normalization from deep SNNs via topology-aware weight standardization folded into static weights, matching dynamic BN accuracy on ImageNet (68.05%) while cutting FPGA neuron LUT usage by 96.4%.
-
A Classifier-Agnostic Zero-Shot Adversarial Attack Detection via CLIP
A^4D detects adversarial attacks in an attack- and classifier-agnostic way by measuring non-arbitrary shifts in CLIP embedding space from prompt-based similarity scores.
-
How to Evaluate and Refine your CAM
Introduces synthetic ground-truth dataset for CAM evaluation, proposes ARCC composite metric, and RefineCAM method that aggregates layers for higher-resolution maps outperforming baselines.
-
PromptDx: Differentiable Prompt Tuning for Multimodal In-Context Alzheimer's Diagnosis
PromptDx adds a differentiable adapter to align multimodal data with a pre-trained TabPFN-style ICL engine, achieving strong Alzheimer's diagnosis performance with only 1% context samples.
-
MoCA3D: Monocular 3D Bounding Box Prediction in the Image Plane
MoCA3D formulates monocular 3D box prediction as dense pixel-space tasks using corner heatmaps and depth maps, with a new PAG metric for image-plane evaluation.
-
TMI: Text-to-Image Meets Image-to-Image for Complementary Data Synthesis to Boost Long-Tailed Instance Segmentation
Hybrid T2I generation with teacher-student pseudo-labeling plus VRAIN context-aware I2I rare-class editing improves LVIS instance segmentation AP, especially on rare categories.
-
AdaBoosting Text Prompts for Vision-Language Models
TPB is an AdaBoost-style ensemble method for text prompts in VLMs that improves few-shot accuracy by targeting hard examples and maintains gains across model transfers.
-
Echo4DIR: 4D Implicit Heart Reconstruction from 2D Echocardiography Videos
Echo4DIR reconstructs continuous 4D cardiac geometry from sparse 2D echocardiography videos using implicit representations, epipolar feature fusion, self-supervised domain adaptation, and radial SDF alignment to achieve up to 98.35% Dice overlap.
-
Do Vision Models Truly Forget? New Findings from Representation-Level Certification of Visual Unlearning in Vertical Federated Learning
Mirage auditing framework reveals that VFL unlearning methods passing output-level certification retain substantial class structure in representations, with no method achieving high utility plus both output and representation forgetting, plus class-sample asymmetry in residual traces.
-
The Learnability Gap in Medical Latent Diffusion
Pretrained autoencoders in medical latent diffusion encode discriminative features well for reconstruction but structure their latent spaces in ways that hinder classifier learning, a gap that persists across architectures and is not closed by domain fine-tuning.
-
Towards Fine-Grained and Verifiable Concept Bottleneck Models
A verifiable CBM framework grounds concepts in localized image patches, achieving comparable accuracy to standard CBMs on medical benchmarks while enabling direct inspection of concept correctness.
-
From Image Hashing to Scene Change Detection
HashSCD is a patch-wise hashing method for unsupervised scene change detection and localization that operates directly in Hamming space with competitive performance and lower computational cost.
-
Counterfactual Stress Testing for Image Classification Models
Counterfactual stress testing with causal generative models offers a more accurate proxy than simple perturbations for predicting medical image model performance under distribution shifts.
-
Hyp2Former: Hierarchy-Aware Hyperbolic Embeddings for Open-Set Panoptic Segmentation
Hyp2Former learns hierarchical semantic similarities in hyperbolic space among known categories so that unknown objects remain close to higher-level concepts and can be detected reliably.
-
Distilling Vision Transformers for Distortion-Robust Representation Learning
An asymmetric multi-level distillation framework lets a student ViT approximate clean-image representations from distorted inputs alone, outperforming prior methods on classification under distortions.
-
Efficient Logic Gate Networks for Video Copy Detection
Logic Gate Networks produce compact Boolean-circuit descriptors for video copy detection that match or exceed prior accuracy at over 11k inferences per second and orders-of-magnitude smaller size.
-
LLM as a Tool, Not an Agent: Code-Mined Tree Transformations for Neural Architecture Search
LLMasTool improves neural architecture search by evolving code-mined hierarchical trees with diversity-guided Bayesian planning and targeted LLM assistance, reporting gains of 0.69, 1.83, and 2.68 points on CIFAR-10, CIFAR-100, and ImageNet16-120.
-
ESCAPE: Episodic Spatial Memory and Adaptive Execution Policy for Long-Horizon Mobile Manipulation
ESCAPE combines spatio-temporal fusion mapping for depth-free 3D memory with a memory-driven grounding module and adaptive execution policy to reach 65.09% success on ALFRED test-seen long-horizon mobile manipulation tasks.
-
A High-Resolution Landscape Dataset for Concept-Based XAI With Application to Species Distribution Models
A new high-resolution landscape concept dataset enables the first Robust TCAV explanations of deep SDMs, largely validating ecological expectations for Plecoptera and Trichoptera while flagging novel associations.
-
GroundingAnomaly: Spatially-Grounded Diffusion for Few-Shot Anomaly Synthesis
GroundingAnomaly uses a Spatial Conditioning Module and Gated Self-Attention in a frozen diffusion U-Net to synthesize spatially accurate few-shot anomalies, reaching SOTA on MVTec AD and VisA for detection, segmentation, and instance detection.
-
PRISM-CTG: A Foundation Model for Cardiotocography Analysis with Multi-View SSL
PRISM-CTG is the first large-scale foundation model for cardiotocography that uses multi-view self-supervised learning on unlabeled data to learn transferable representations, outperforming baselines on seven downstream tasks with external validation.
-
Clinical Cognition Alignment for Gastrointestinal Diagnosis with Multimodal LLMs
Hierarchical clinical-reasoning SFT plus counterfactual GRPO yields SoTA diagnostic accuracy for multimodal LLMs on gastrointestinal endoscopy benchmarks.
-
Learning to Unify Deformable Shape and Texture Representations for Cardiac Video Classification
ShapeFuse uses bidirectional cross-modal temporal attention and adaptive gating to fuse deformable shape and texture features for cardiac video classification, outperforming existing fusion strategies on a cine CMR dataset.
-
FrameONE: Hierarchical Motion Modeling for Universal Multi-View Echocardiographic Keyframe Detection
FrameONE introduces hierarchical intra-view and inter-view motion modeling for end-to-end multi-view echocardiographic keyframe detection, reporting SOTA accuracy on 25,872 videos from four views with code released.
-
HyperVLP: Enhancing Hierarchical Surgical Video-Language Pre-training in Hyperbolic Space
HyperVLP uses hyperbolic geometry in surgical video-language pre-training to preserve hierarchy across actions, steps, and phases, yielding gains in zero- and few-shot phase recognition.
-
Fisher-Routed Mixture of Experts for Federated Class-Incremental Learning
FedFMX adds Fisher-routed expert selection and routing-aware regularization to federated class-incremental learning and proves an O(T^{-1}) convergence rate.
-
Distill Once, Adapt Life-Long: Exploring Dataset Distillation for Continual Test-Time Adaptation
DO-ALL applies dataset distillation to generate synthetic source anchors that stabilize continual test-time adaptation under evolving domains without storing original source data.
-
Worst-Group Equalized Odds Regularization for Multi-Attribute Fair Medical Image Classification
A worst-group equalized odds regularizer targets extreme subgroup deviations in true and false positive rates to improve multi-attribute fairness in medical imaging while preserving AUC.
-
Road Maps as Free Geometric Priors: Weather-Invariant Drone Geo-Localization with GeoFuse
GeoFuse fuses aligned road maps with satellite imagery via token/channel interactions and dynamic gating plus contrastive learning, lifting Recall@1 by 3.46% on University-1652 and 23.18% on DenseUAV under weather degradation.
-
Hybrid Quantum-MambaVision: A Quantum-enhanced State Space Model for Calibrated Mixed-type Wafer Defect Detection
Hybrid Quantum-MambaVision combines Mamba SSM with a parameterized quantum adapter for improved multi-label classification and calibration on the imbalanced MixedWM38 wafer defect dataset.
-
FedSurrogate: Backdoor Defense in Federated Learning via Layer Criticality and Surrogate Replacement
FedSurrogate defends federated learning against backdoors by clustering on security-critical layers and substituting malicious updates with benign surrogates, reporting false-positive rates below 10% and attack success below 2.1% under non-IID conditions.
-
Sketch and Text Synergy: Fusing Structural Contours and Descriptive Attributes for Fine-Grained Image Retrieval
STBIR fuses sketches and text via curriculum robustness, category optimization, and staged alignment to outperform prior methods on a new fine-grained benchmark dataset.
-
QShield: Securing Neural Networks Against Adversarial Attacks using Quantum Circuits
Hybrid quantum-classical models using structured entanglement keep high accuracy on MNIST, OrganAMNIST and CIFAR-10 while lowering adversarial attack success rates and raising the computational cost of generating attacks.
-
DBMF: A Dual-Branch Multimodal Framework for Out-of-Distribution Detection
DBMF integrates scores from text-image and vision branches to improve out-of-distribution detection on endoscopic datasets by up to 24.84% over prior methods.
-
Accuracy Improvement of Cell Image Segmentation Using Feedback Former
Feedback Former improves cell image segmentation accuracy by feeding detailed feature maps back from near the output to lower transformer layers, outperforming non-feedback baselines with lower computational cost on three datasets.
-
Toxicity Prediction by Multimodal Deep Learning
A multimodal deep learning approach using heterogeneous representations and network types achieves significantly higher accuracy than state-of-the-art methods on a standard toxicity benchmark.
-
Learning Coarse-to-Fine Osteoarthritis Representations under Noisy Hierarchical Labels
Dual-head training on hierarchical OA labels yields backbone-dependent gains in KL metrics, more ordered latent severity axes, and better saliency alignment with cartilage for some 3D backbones.
-
STAND: Semantic Anchoring Constraint with Dual-Granularity Disambiguation for Remote Sensing Image Change Captioning
STAND adds semantic anchoring and dual-granularity disambiguation modules to address viewpoint, scale, and knowledge ambiguities in remote sensing change captioning.
-
On the Properties of Feature Attribution for Supervised Contrastive Learning
Neural networks trained via supervised contrastive learning yield feature attributions that are more faithful, less complex, and more continuous than those from cross-entropy trained networks.
-
MsEdF: A Multi-stream Encoder-decoder Framework for Remote Sensing Image Captioning
MsEdF combines two complementary image encoders for feature diversity and a stacked GRU decoder with element-wise aggregation to improve remote sensing image captioning on three benchmark datasets.
-
Generalized SAM: Efficient Fine-Tuning of SAM for Variable Input Image Sizes
GSAM applies random cropping to enable variable input sizes for efficient SAM fine-tuning, claiming lower compute with comparable or higher accuracy on varied datasets.
-
Visual graphs for image classification: does the structure affect performance?
Sparse graph topologies (especially minimum spanning trees) match or exceed denser graphs for image classification on Fashion-MNIST when using a fixed three-layer GCN.
-
DMDSC: A Dynamic-Margin Deep Simplex Classifier for Open-Set Recognition on Medical Image Datasets
DMDSC adapts class-specific margins dynamically by label frequency in deep simplex classifiers to improve open-set recognition on imbalanced medical image datasets.