Retrieval from motion datasets combined with LLM task parsing and reward-guided noise initialization enables training-free diffusion optimization to satisfy severe spatiotemporal constraints in human motion generation.
hub
Not all noises are created equally: Diffusion noise selection and optimization
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
hub tools
citation-role summary
background 3
citation-polarity summary
roles
background 3polarities
background 3representative citing papers
Oracle Noise optimizes diffusion model noise on a Riemannian hypersphere guided by key prompt words to preserve the Gaussian prior, eliminate norm inflation, and achieve faster semantic alignment than Euclidean methods.
VideoRepair detects text-video misalignments via MLLM-generated questions and performs localized, region-preserving refinement to improve alignment in existing T2V diffusion models.
PATHS applies parallel tempering to improve initial particle sampling for SMC reward alignment, yielding better results on layout-to-image and quantity-aware generation tasks.
ABSS ranks diffusion seeds by early cross-attention strength to prompt core tokens and retains only the top-k for full generation, yielding consistent gains in alignment and quality on Stable Diffusion variants.
Coupled initial noises in diffusion models, with designed dependence but unchanged marginal Gaussians, improve generated image diversity on Stable Diffusion variants while preserving quality and alignment.
LIVEditor-14B applies a new sparse attention method (ISA) that prunes context and uses query-sharpness routing to cut attention latency ~60% with no loss in editing quality on standard benchmarks.
NoiseRater meta-learns instance-level importance scores for noise in diffusion training via bilevel optimization, then uses a two-stage pipeline to improve efficiency and generation quality on FFHQ and ImageNet.
A single constant initial noise vector found by Monte-Carlo search improves frozen diffusion and flow-matching robot policies on 46 of 51 tasks, by up to 55% absolute success rate.
Diffusion models improve generation quality via inference-time search over noise candidates guided by verifiers and algorithms, yielding gains beyond denoising step scaling on class- and text-conditioned benchmarks.
FDM-MFVT is a few-step mask-free virtual try-on diffusion model using OANO and IDT modules plus a new 30,000-pair MFVT dataset, claiming better efficiency and quality than baselines.
TapSampling improves generalist robotic manipulation policies at inference time via latent action sampling with an Action-VAE and selection by a task-progress outcome predictor.
citing papers explorer
-
Towards Highly-Constrained Human Motion Generation with Retrieval-Guided Diffusion Noise Optimization
Retrieval from motion datasets combined with LLM task parsing and reward-guided noise initialization enables training-free diffusion optimization to satisfy severe spatiotemporal constraints in human motion generation.
-
Oracle Noise: Faster Semantic Spherical Alignment for Interpretable Latent Optimization
Oracle Noise optimizes diffusion model noise on a Riemannian hypersphere guided by key prompt words to preserve the Gaussian prior, eliminate norm inflation, and achieve faster semantic alignment than Euclidean methods.
-
Self-Correcting Text-to-Video Generation with Misalignment Detection and Localized Refinement
VideoRepair detects text-video misalignments via MLLM-generated questions and performs localized, region-preserving refinement to improve alignment in existing T2V diffusion models.
-
Parallel Tempering Initial Sampling in Inference-Time Reward Alignment
PATHS applies parallel tempering to improve initial particle sampling for SMC reward alignment, yielding better results on layout-to-image and quantity-aware generation tasks.
-
Boosting Text-to-Image Diffusion Models via Core Token Attention-Based Seed Selection
ABSS ranks diffusion seeds by early cross-attention strength to prompt core tokens and retains only the top-k for full generation, yielding consistent gains in alignment and quality on Stable Diffusion variants.
-
Couple to Control: Joint Initial Noise Design in Diffusion Models
Coupled initial noises in diffusion models, with designed dependence but unchanged marginal Gaussians, improve generated image diversity on Stable Diffusion variants while preserving quality and alignment.
-
LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention
LIVEditor-14B applies a new sparse attention method (ISA) that prunes context and uses query-sharpness routing to cut attention latency ~60% with no loss in editing quality on standard benchmarks.
-
NoiseRater: Meta-Learned Noise Valuation for Diffusion Model Training
NoiseRater meta-learns instance-level importance scores for noise in diffusion training via bilevel optimization, then uses a two-stage pipeline to improve efficiency and generation quality on FFHQ and ImageNet.
-
You've Got a Golden Ticket: Improving Generative Robot Policies With A Single Noise Vector
A single constant initial noise vector found by Monte-Carlo search improves frozen diffusion and flow-matching robot policies on 46 of 51 tasks, by up to 55% absolute success rate.
-
Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Steps
Diffusion models improve generation quality via inference-time search over noise candidates guided by verifiers and algorithms, yielding gains beyond denoising step scaling on class- and text-conditioned benchmarks.
-
FDM-MFVT: Few-step Sampling Diffusion Model for Mask-Free Virtual Try-On
FDM-MFVT is a few-step mask-free virtual try-on diffusion model using OANO and IDT modules plus a new 30,000-pair MFVT dataset, claiming better efficiency and quality than baselines.
-
TapSampling: Inference-Time Sampling with a Task-Progress-Understanding Verifier for Robotic Manipulation
TapSampling improves generalist robotic manipulation policies at inference time via latent action sampling with an Action-VAE and selection by a task-progress outcome predictor.