CDM migrates distribution matching distillation to continuous time via dynamic random-length schedules and active off-trajectory latent alignment, yielding competitive few-step image fidelity on SD3 and Longcat-Image.
hub Tool reference
Microsoft coco: Common objects in context
Tool reference. 78% of classified Pith citations use this work as a method, library, or software dependency, not as a substantive claim.
29
Pith papers citing it
Method reference
78% of classified citations
hub tools
citation-role summary
dataset 6
background 2
method 1
citation-polarity summary
representative citing papers
Introduces Synergistic Faithfulness metric based on Shapley Interaction Index to evaluate cross-modal synergy in VLM explainers, revealing over-reliance on visual salience in existing methods.
HeadKV compresses KV cache for autoregressive image generation via head-aware budget allocation, early head-type identification from consistent patterns, and stratified token eviction.
3DEditSafe adds generation-stage guidance, 3D safety regularization, semantic projection, residue suppression, and mask-aware preservation to reduce unsafe semantic alignment in 3D editing while noting a safety-quality tradeoff.
PODS is a plug-and-play oscillatory data-volume scheduler that alternates low-ratio regularization phases with high-ratio recovery phases to improve data selection efficiency across training tasks.
SAEParate disentangles sparse representations in diffusion models via contrastive clustering and nonlinear encoding to enable more precise concept unlearning with reduced side effects.
DistractMIA performs output-only black-box membership inference on vision-language models by inserting semantic distractors and measuring shifts in generated text responses.
TCP-SSM conditions stable poles on visual tokens to explicitly control memory decay and oscillation in SSMs, cutting computation up to 44% while matching or exceeding accuracy on classification, segmentation, and detection.
PoseBridge recovers semantic information lost during skeletonization by extracting pose-anchored cues from human pose estimation and transferring them via skeleton-conditioned bridging and semantic prototype adaptation, yielding 13.3-17.4 point gains on the Kinetics PURLS benchmark.
CAFE benchmark reveals that promptable segmentation models often produce correct masks for misleading prompts, showing a gap between localization accuracy and true concept understanding.
SAM 3D Animal is the first promptable framework for multi-animal 3D reconstruction from single images, built on SMAL+ and trained on the new Herd3D dataset, achieving SOTA results on Animal3D, APTv2, and Animal Kingdom benchmarks.
InstructMoLE replaces per-token routing with instruction-guided global routing for mixture-of-low-rank-experts in diffusion transformers and adds an output-space orthogonality loss to improve multi-conditional image generation.
CortexMAE adapts Vision Transformers to fMRI via cortical flat maps, shows power-law scaling on 2.1K hours of data, and outperforms priors on cognitive state decoding while failing to beat a simple functional connectivity baseline on subject-level trait prediction.
LAION-5B is an openly released dataset of 5.85 billion CLIP-filtered image-text pairs that enables replication of foundational vision-language models.
DarkLLM trains an LLM to generate language-driven adversarial perturbations that unify targeted, untargeted, segmentation, and multi-model attacks on foundation models.
Method converts exocentric videos to egocentric format via body-pose extraction and kinematics to improve egocentric world-model prediction and planning.
Nonlinear Bipolar Compensation with Bipolar Logarithmic Transformation reduces outlier effects in post-training quantization by performing compensation in a compressed transformed space.
Pretrained instruction-based image editing models exhibit early foreground-background separability that enables a training-free framework for zero-shot referring image segmentation using a single denoising step.
A unified visual conditioning approach fuses semantic and appearance features before VLM processing, with two-stage training and slot-wise regularization, to improve consistency in multi-reference image generation.
LIME reduces hallucinations in multimodal LLMs by using LRP to boost perceptual modality contributions through inference-time KV updates.
Dynamic parameterization of standard layers can replace explicit attention for linear-time global visual modeling.
SLQ adapts frozen MLLMs for multimodal retrieval by appending shared latent queries to text and image tokens and introduces KARR-Bench to test knowledge-aware reasoning retrieval.
Rosetta proposes a composable multimodal pretraining method with MAOP to prevent catastrophic forgetting when expanding modalities beyond standard MoE and MoT approaches.
TinyFormer adds Parallel Bi-fusion Module and Spatial Semantic Adapter to a YOLO-DETR hybrid, raising small-object AP by 1.6 points to 58.5% on MS COCO while keeping real-time speed.
citing papers explorer
-
Continuous-Time Distribution Matching for Few-Step Diffusion Distillation
CDM migrates distribution matching distillation to continuous time via dynamic random-length schedules and active off-trajectory latent alignment, yielding competitive few-step image fidelity on SD3 and Longcat-Image.
-
Measuring Cross-Modal Synergy: A Benchmark for VLM Explainability
Introduces Synergistic Faithfulness metric based on Shapley Interaction Index to evaluate cross-modal synergy in VLM explainers, revealing over-reliance on visual salience in existing methods.
-
Head-Aware Key-Value Compression for Efficient Autoregressive Image Generation
HeadKV compresses KV cache for autoregressive image generation via head-aware budget allocation, early head-type identification from consistent patterns, and stratified token eviction.
-
3DEditSafe: Defending 3D Editing Pipelines from Unsafe Generation
3DEditSafe adds generation-stage guidance, 3D safety regularization, semantic projection, residue suppression, and mask-aware preservation to reduce unsafe semantic alignment in 3D editing while noting a safety-quality tradeoff.
-
Beyond What to Select: A Plug-and-play Oscillatory Data-Volume Scheduling for Efficient Model Training
PODS is a plug-and-play oscillatory data-volume scheduler that alternates low-ratio regularization phases with high-ratio recovery phases to improve data selection efficiency across training tasks.
-
Disentangled Sparse Representations for Concept-Separated Diffusion Unlearning
SAEParate disentangles sparse representations in diffusion models via contrastive clustering and nonlinear encoding to enable more precise concept unlearning with reduced side effects.
-
DistractMIA: Black-Box Membership Inference on Vision-Language Models via Semantic Distraction
DistractMIA performs output-only black-box membership inference on vision-language models by inserting semantic distractors and measuring shifts in generated text responses.
-
TCP-SSM: Efficient Vision State Space Models with Token-Conditioned Poles
TCP-SSM conditions stable poles on visual tokens to explicitly control memory decay and oscillation in SSMs, cutting computation up to 44% while matching or exceeding accuracy on classification, segmentation, and detection.
-
PoseBridge: Bridging the Skeletonization Gap for Zero-Shot Skeleton-Based Action Recognition
PoseBridge recovers semantic information lost during skeletonization by extracting pose-anchored cues from human pose estimation and transferring them via skeleton-conditioned bridging and semantic prototype adaptation, yielding 13.3-17.4 point gains on the Kinetics PURLS benchmark.
-
From Pixels to Concepts: Do Segmentation Models Understand What They Segment?
CAFE benchmark reveals that promptable segmentation models often produce correct masks for misleading prompts, showing a gap between localization accuracy and true concept understanding.
-
SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild
SAM 3D Animal is the first promptable framework for multi-animal 3D reconstruction from single images, built on SMAL+ and trained on the new Herd3D dataset, achieving SOTA results on Animal3D, APTv2, and Animal Kingdom benchmarks.
-
InstructMoLE: Instruction-Guided Mixture of Low-rank Experts for Multi-Conditional Image Generation
InstructMoLE replaces per-token routing with instruction-guided global routing for mixture-of-low-rank-experts in diffusion transformers and adds an output-space orthogonality loss to improve multi-conditional image generation.
-
Scaling Vision Transformers for Functional MRI with Flat Maps
CortexMAE adapts Vision Transformers to fMRI via cortical flat maps, shows power-law scaling on 2.1K hours of data, and outperforms priors on cognitive state decoding while failing to beat a simple functional connectivity baseline on subject-level trait prediction.
-
LAION-5B: An open large-scale dataset for training next generation image-text models
LAION-5B is an openly released dataset of 5.85 billion CLIP-filtered image-text pairs that enables replication of foundational vision-language models.
-
DarkLLM: Learning Language-Driven Adversarial Attacks with Large Language Models
DarkLLM trains an LLM to generate language-driven adversarial perturbations that unify targeted, untargeted, segmentation, and multi-model attacks on foundation models.
-
EgoExo-WM: Unlocking Exo Video for Ego World Models
Method converts exocentric videos to egocentric format via body-pose extraction and kinematics to improve egocentric world-model prediction and planning.
-
Nonlinear Bipolar Compensation: Handling Outliers in Post-Training Quantization
Nonlinear Bipolar Compensation with Bipolar Logarithmic Transformation reduces outlier effects in post-training quantization by performing compensation in a compressed transformed space.
-
Early Semantic Grounding in Image Editing Models for Zero-Shot Referring Image Segmentation
Pretrained instruction-based image editing models exhibit early foreground-background separability that enables a training-free framework for zero-shot referring image segmentation using a single denoising step.
-
UniCustom: Unified Visual Conditioning for Multi-Reference Image Generation
A unified visual conditioning approach fuses semantic and appearance features before VLM processing, with two-stage training and slot-wise regularization, to improve consistency in multi-reference image generation.
-
Mitigating Multimodal LLMs Hallucinations via Relevance Propagation at Inference Time
LIME reduces hallucinations in multimodal LLMs by using LRP to boost perceptual modality contributions through inference-time KV updates.
-
Linear-Time Global Visual Modeling without Explicit Attention
Dynamic parameterization of standard layers can replace explicit attention for linear-time global visual modeling.
-
SLQ: Bridging Modalities via Shared Latent Queries for Retrieval with Frozen MLLMs
SLQ adapts frozen MLLMs for multimodal retrieval by appending shared latent queries to text and image tokens and introduces KARR-Bench to test knowledge-aware reasoning retrieval.
-
Rosetta: Composable Native Multimodal Pretraining
Rosetta proposes a composable multimodal pretraining method with MAOP to prevent catastrophic forgetting when expanding modalities beyond standard MoE and MoT approaches.
-
TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid Real-time Detectors
TinyFormer adds Parallel Bi-fusion Module and Spatial Semantic Adapter to a YOLO-DETR hybrid, raising small-object AP by 1.6 points to 58.5% on MS COCO while keeping real-time speed.
-
Frequency-Domain Regularized Adversarial Alignment for Transferable Attacks against Closed-Source MLLMs
FRA-Attack uses high-pass DCT feature alignment and frequency-domain gradient regularization to boost adversarial transferability across 15 MLLMs from 7 vendors.
-
CAST: Collapse-Aware multi-Scale Topology Fusion for Multimodal Coreset Selection
CAST selects better multimodal coresets by fusing collapse-aware topologies across modalities and matching distributions at multiple scales in the diffusion wavelet domain.
-
InsHuman: Towards Natural and Identity-Preserving Human Insertion
InsHuman proposes Human-Background Adaptive Fusion, Face-to-Face ID-Preserving, and Bidirectional Data Pairing to enable natural human insertion in images without altering identity.
-
VL-SAM-v3: Memory-Guided Visual Priors for Open-World Object Detection
VL-SAM-v3 retrieves visual prototypes from memory to generate sparse spatial and dense contextual priors that refine detection prompts, yielding gains on rare categories in LVIS for both open-vocabulary and open-ended settings.
-
Colinearity Decay: Training Quantization-Friendly ViTs with Outlier Decay
Colinearity-Decay regularizer trains ViTs that maintain or improve full-precision accuracy while delivering higher accuracy after low-bit quantization on ImageNet and COCO tasks.