LumaFlux: Lifting 8-Bit Worlds to HDR Reality with Physically-Guided Diffusion Transformers
Paper • 2604.02787 • Published
Diffusion Research
Research work on Guidance and test-time search for diffusion models, efficient sampling, post-training, RL, perceptual quality,HDR-UGC video quality,
-
-
Seeing Beyond 8bits: Subjective and Objective Quality Assessment of HDR-UGC Videos
Paper • 2603.00938 • Published -
GameScope: A Multi-Attribute, Multi-Codec Benchmark Dataset for Gaming Video Quality Assessment
Paper • 2605.01272 • Published -
CHUG: Crowdsourced User-Generated HDR Video Quality Dataset
Paper • 2510.09879 • Published -
shreshthsaini/Beyond8Bits
Viewer • Updated • 47.4k • 41
shreshthsaini/BrightVQ
Viewer • Updated • 2.1k • 25Note Dataset for BrightRate: Quality Assessment for User-Generated HDR Videos (WACV 2026, oral, Best Student Paper Award). Paper: https://openaccess.thecvf.com/content/WACV2026/html/Saini_BrightRate_Quality_Assessment_for_User-Generated_HDR_Videos_WACV_2026_paper.html Code: https://github.com/shreshthsaini/BrightVQ
-
shreshthsaini/CHUG
Viewer • Updated • 5.99k • 23 -
shreshthsaini/LumaFlux
Image-to-Image • Updated -
shreshthsaini/brightrate-lm-7b-sdr
Image-Text-to-Text • Updated • 9 -
shreshthsaini/brightrate-lm-7b-multiexposure
Image-Text-to-Text • Updated • 10 -
CachedSearch: Training-Free Cached Exploration for Test-Time Search in Video Diffusion
Paper • 2607.23159 • Published -
Rectified-CFG++ for Flow Based Models
Paper • 2510.07631 • Published -
LumaGuide: Distribution Shaping for Training-Free HDR Generation in Diffusion Models
Paper • 2607.26237 • Published -
Latent Guidance in Diffusion Models for Perceptual Evaluations
Paper • 2506.00327 • Published -
HIDRO-VQA: High Dynamic Range Oracle for Video Quality Assessment
Paper • 2311.11059 • Published -
shreshthsaini/LumaGuide
Text-to-Image • Updated
MEND: RL For Flow Models via Proximal Velocity Matching
Paper • 2610.05954 • Published • 3Note MEND: RL for Flow Models via Proximal Velocity Matching. Official code, weights, project page and blog: https://github.com/shreshthsaini/MEND-RL ; models: https://huggingface.co/collections/shreshthsaini/mend-rl-for-flow-models-via-proximal-velocity-matching-6ac5d201a2821178270f97ce
-
JLD: Perceptual Distance Through A Jacobian Lens
Paper • 2610.05967 • Published • 4 -
TABES: Trajectory-Aware Backward-on-Entropy Steering for Masked Diffusion Models
Paper • 2602.00250 • Published -
shreshthsaini/Rectified-CFGpp
Text-to-Image • Updated -
shreshthsaini/CachedSearch
Text-to-Video • Updated • 1 -
shreshthsaini/JLD
Other • Updated • 35 • 1