ReConText3D is the first replay-memory framework for continual text-to-3D generation that prevents catastrophic forgetting on new textual categories while preserving quality on previously seen classes.
hub Canonical reference
High-resolution image synthesis with latent diffusion models
Canonical reference. 100% of citing Pith papers cite this work as background.
23
Pith papers citing it
Background
100% of classified citations
hub tools
citation-role summary
background 6
citation-polarity summary
roles
background 6polarities
background 6representative citing papers
OrbitQuant is a data-agnostic PTQ technique for DiTs that uses RPBH rotation in a normalized basis to enable a single codebook across all inputs, achieving SOTA low-bit performance on FLUX.1, CogVideoX and similar models.
VDE accelerates rectified flow models like Flux by 3.22x with LPIPS of 0.069 via velocity decomposition into parallel/orthogonal components plus periodic full-pass anchoring.
CoReDiT reduces self-attention FLOPs in DiTs by up to 55% via linear-time spatial coherence pruning and neighbor-based reconstruction, delivering 1.33x-1.72x speedups with maintained quality.
Uni-AdGen uses a unified autoregressive framework with foreground perception, instruction tuning, and coarse-to-fine preference modules to generate personalized image-text ads from noisy user behaviors, outperforming baselines on a new PAd1M dataset.
Omni-NegCLIP improves CLIP's negation understanding by up to 52.65% on presence-based and 12.50% on absence-based tasks through front-layer fine-tuning with specialized contrastive losses.
UniDAC achieves universal metric depth estimation across camera types by decoupling relative depth prediction from spatially varying scale estimation using a depth-guided module and distortion-aware positional embedding.
Stroke of Surprise is a framework that generates vector sketches undergoing semantic transformation from one concept to another by adding strokes, using dual-branch SDS and overlay loss for optimization.
CoMoVi co-generates 3D human motions and 2D videos synchronously in a single diffusion denoising loop using 3D-to-2D projection and dual-branch diffusion with 3D-2D cross attentions.
A two-stage framework that predicts a Canny edge map before rendering improves preservation of high-frequency identity details (logos, patterns, text) in subject-driven image generation.
AttriStory adds a benchmark and AttriLoss-based latent optimization to improve faithful rendering of fine-grained attributes such as clothing color and texture in diffusion-model visual storytelling.
Conditional score-based diffusion models synthesize phase maps from magnitude-only MR images to generate k-space datasets that train superior deep learning models for accelerated MRI reconstruction compared to smooth-phase or GAN-based alternatives.
PDF-GS progressively filters distractors in 3D Gaussian Splatting by exploiting the method's self-suppression of inconsistent signals, yielding high-fidelity distractor-free 3D models without changing the base architecture or adding inference cost.
VAG is a synchronized dual-stream flow-matching framework that generates aligned video-action pairs for synthetic embodied data synthesis and policy pretraining.
Position and step penalty plus visual reasoning guidance fix premature answering and weak visual grounding in diffusion MLLMs, delivering up to 7.5% accuracy gains and over 3x speedup.
Unlearning methods that strongly erase concepts from text-to-image diffusion models consistently degrade performance on attribute binding, spatial reasoning, and counting tasks.
MaskDiME uses adaptive masked diffusion to produce 30x faster, localized, and semantically consistent visual counterfactual explanations without training, matching or exceeding prior performance on five datasets.
MEBench is a new benchmark and data-generation pipeline that measures mutual exclusivity bias in VLMs, finding weak bias but some use of spatial context to resolve novel-object ambiguity.
RealDiffusion uses heat diffusion as a dissipative prior and a region-aware stochastic process inside a training-free physics-informed attention mechanism to improve multi-character coherence while preserving narrative dynamism in sequential image generation.
Unposed-to-3D learns simulation-ready 3D vehicle models from unposed real images by predicting camera parameters for photometric self-supervision, then adding scale prediction and harmonization.
A scalable pipeline generates an intra-consistent, inter-diverse 1.4M style image dataset from text-to-image models and uses it to train a style encoder and generalizable style transfer model.
A framework that combines MLLM-based image enhancement with a medium-aware 3D Gaussian Splatting model to reconstruct and render smoke scenes.
Two new lightweight modules for diffusion-based real-world image super-resolution deliver competitive perceptual quality and better structure preservation on DIV2K and RealSR datasets.
citing papers explorer
-
ReConText3D: Replay-based Continual Text-to-3D Generation
ReConText3D is the first replay-memory framework for continual text-to-3D generation that prevents catastrophic forgetting on new textual categories while preserving quality on previously seen classes.
-
OrbitQuant: Data-Agnostic Quantization for Image and Video Diffusion Transformers
OrbitQuant is a data-agnostic PTQ technique for DiTs that uses RPBH rotation in a normalized basis to enable a single codebook across all inputs, achieving SOTA low-bit performance on FLUX.1, CogVideoX and similar models.
-
VDE: Training-Free Accelerating Rectified Flow Model via Velocity Decomposition and Estimation
VDE accelerates rectified flow models like Flux by 3.22x with LPIPS of 0.069 via velocity decomposition into parallel/orthogonal components plus periodic full-pass anchoring.
-
CoReDiT: Spatial Coherence-Guided Token Pruning and Reconstruction for Efficient Diffusion Transformers
CoReDiT reduces self-attention FLOPs in DiTs by up to 55% via linear-time spatial coherence pruning and neighbor-based reconstruction, delivering 1.33x-1.72x speedups with maintained quality.
-
Design Your Ad: Personalized Advertising Image and Text Generation with Unified Autoregressive Models
Uni-AdGen uses a unified autoregressive framework with foreground perception, instruction tuning, and coarse-to-fine preference modules to generate personalized image-text ads from noisy user behaviors, outperforming baselines on a new PAd1M dataset.
-
Omni-NegCLIP: Enhancing CLIP with Front-Layer Contrastive Fine-Tuning for Comprehensive Negation Understanding
Omni-NegCLIP improves CLIP's negation understanding by up to 52.65% on presence-based and 12.50% on absence-based tasks through front-layer fine-tuning with specialized contrastive losses.
-
UniDAC: Universal Metric Depth Estimation for Any Camera
UniDAC achieves universal metric depth estimation across camera types by decoupling relative depth prediction from spatially varying scale estimation using a depth-guided module and distortion-aware positional embedding.
-
Stroke of Surprise: Progressive Semantic Illusions in Vector Sketching
Stroke of Surprise is a framework that generates vector sketches undergoing semantic transformation from one concept to another by adding strokes, using dual-branch SDS and overlay loss for optimization.
-
CoMoVi: Co-Generation of 3D Human Motions and Realistic Videos
CoMoVi co-generates 3D human motions and 2D videos synchronously in a single diffusion denoising loop using 3D-to-2D projection and dual-branch diffusion with 3D-2D cross attentions.
-
Decomposing Subject-Driven Image Generation via Intermediate Structural Prediction
A two-stage framework that predicts a Canny edge map before rendering improves preservation of high-frequency identity details (logos, patterns, text) in subject-driven image generation.
-
AttriStory: Fine-grained Attribute Realization for Visual Storytelling with Diffusion Models
AttriStory adds a benchmark and AttriLoss-based latent optimization to improve faithful rendering of fine-grained attributes such as clothing color and texture in diffusion-model visual storytelling.
-
Phase-map synthesis from magnitude-only MR images using conditional score-based diffusion models with application in training of accelerated MRI reconstruction models
Conditional score-based diffusion models synthesize phase maps from magnitude-only MR images to generate k-space datasets that train superior deep learning models for accelerated MRI reconstruction compared to smooth-phase or GAN-based alternatives.
-
PDF-GS: Progressive Distractor Filtering for Robust 3D Gaussian Splatting
PDF-GS progressively filters distractors in 3D Gaussian Splatting by exploiting the method's self-suppression of inconsistent signals, yielding high-fidelity distractor-free 3D models without changing the base architecture or adding inference cost.
-
VAG: Dual-Stream Video-Action Generation for Embodied Data Synthesis
VAG is a synchronized dual-stream flow-matching framework that generates aligned video-action pairs for synthetic embodied data synthesis and policy pretraining.
-
Thinking Diffusion: Penalize and Guide Visual-Grounded Reasoning in Diffusion Multimodal Language Models
Position and step penalty plus visual reasoning guidance fix premature answering and weak visual grounding in diffusion MLLMs, delivering up to 7.5% accuracy gains and over 3x speedup.
-
Erasure or Erosion? Evaluating Compositional Degradation in Unlearned Text-To-Image Diffusion Models
Unlearning methods that strongly erase concepts from text-to-image diffusion models consistently degrade performance on attribute binding, spatial reasoning, and counting tasks.
-
MaskDiME: Adaptive Masked Diffusion for Precise and Efficient Visual Counterfactual Explanations
MaskDiME uses adaptive masked diffusion to produce 30x faster, localized, and semantically consistent visual counterfactual explanations without training, matching or exceeding prior performance on five datasets.
-
MEBench: A Novel Benchmark for Understanding Mutual Exclusivity Bias in Vision-Language Models
MEBench is a new benchmark and data-generation pipeline that measures mutual exclusivity bias in VLMs, finding weak bias but some use of spatial context to resolve novel-object ambiguity.
-
RealDiffusion: Physics-informed Attention for Multi-character Storybook Generation
RealDiffusion uses heat diffusion as a dissipative prior and a region-aware stochastic process inside a training-free physics-informed attention mechanism to improve multi-character coherence while preserving narrative dynamism in sequential image generation.
-
Unposed-to-3D: Learning Simulation-Ready Vehicles from Real-World Images
Unposed-to-3D learns simulation-ready 3D vehicle models from unposed real images by predicting camera parameters for photometric self-supervision, then adding scale prediction and harmonization.
-
MegaStyle: Constructing Diverse and Scalable Style Dataset via Consistent Text-to-Image Style Mapping
A scalable pipeline generates an intra-consistent, inter-diverse 1.4M style image dataset from text-to-image models and uses it to train a style encoder and generalizable style transfer model.
-
3D Smoke Scene Reconstruction Guided by Vision Priors from Multimodal Large Language Models
A framework that combines MLLM-based image enhancement with a medium-aware 3D Gaussian Splatting model to reconstruct and render smoke scenes.
-
Degradation-Aware and Structure-Preserving Diffusion for Real-World Image Super-Resolution
Two new lightweight modules for diffusion-based real-world image super-resolution deliver competitive perceptual quality and better structure preservation on DIV2K and RealSR datasets.