{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T05:15:25Z","timestamp":1778130925844,"version":"3.51.4"},"reference-count":76,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. on Image Process."],"published-print":{"date-parts":[[2026]]},"DOI":"10.1109\/tip.2026.3684401","type":"journal-article","created":{"date-parts":[[2026,4,21]],"date-time":"2026-04-21T19:51:35Z","timestamp":1776801095000},"page":"4571-4586","source":"Crossref","is-referenced-by-count":0,"title":["RSTFA: Efficient Training-Free Human-Preference Alignment via Rejection Sampling for Text-to-Image Diffusion Models"],"prefix":"10.1109","volume":"35","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-2628-5916","authenticated-orcid":false,"given":"Hongzheng","family":"Yang","sequence":"first","affiliation":[{"name":"The Chinese University of Hong Kong, Hong Kong, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9389-4095","authenticated-orcid":false,"given":"Jason Chun-Lok","family":"Li","sequence":"additional","affiliation":[{"name":"University of Hong Kong, Hong Kong, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-2944-3501","authenticated-orcid":false,"given":"Kun","family":"Li","sequence":"additional","affiliation":[{"name":"City University of Hong Kong, Hong Kong, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9952-5266","authenticated-orcid":false,"given":"Wenao","family":"Ma","sequence":"additional","affiliation":[{"name":"The Chinese University of Hong Kong, Hong Kong, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Mingjie","family":"Xu","sequence":"additional","affiliation":[{"name":"Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8561-2206","authenticated-orcid":false,"given":"Yuzhi","family":"Zhao","sequence":"additional","affiliation":[{"name":"City University of Hong Kong, Hong Kong, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5185-1492","authenticated-orcid":false,"given":"Lai-Man","family":"Po","sequence":"additional","affiliation":[{"name":"City University of Hong Kong, Hong Kong, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73411-3_5"},{"key":"ref2","article-title":"Fast timing-conditioned latent audio diffusion","author":"Evans","year":"2024","journal-title":"arXiv:2402.04825"},{"key":"ref3","first-page":"8780","article-title":"Diffusion models beat GANs on image synthesis","volume-title":"Proc. NIPS","volume":"34","author":"Dhariwal"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2923"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00457"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1038\/s43588-024-00737-x"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1111\/cgf.15063"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/VR58804.2024.00039"},{"key":"ref9","article-title":"InstructPix2NeRF: Instructed 3D portrait editing from a single image","author":"Li","year":"2023","journal-title":"arXiv:2311.02826"},{"key":"ref10","article-title":"Prompt-to-prompt image editing with cross attention control","author":"Hertz","year":"2022","journal-title":"arXiv:2208.01626"},{"key":"ref11","article-title":"Directly fine-tuning diffusion models on differentiable rewards","volume-title":"Proc. 12th Int. Conf. Learn. Represent. (ICLR)","author":"Clark"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00669"},{"key":"ref13","article-title":"Aligning text-to-image diffusion models with reward backpropagation","author":"Prabhudesai","year":"2023","journal-title":"arXiv:2310.03739"},{"key":"ref14","first-page":"4299","article-title":"Deep reinforcement learning from human preferences","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Christiano"},{"key":"ref15","first-page":"53728","article-title":"Direct preference optimization: Your language model is secretly a reward model","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Rafailov"},{"key":"ref16","article-title":"Training a helpful and harmless assistant with reinforcement learning from human feedback","author":"Bai","year":"2022","journal-title":"arXiv:2204.05862"},{"key":"ref17","article-title":"Constitutional AI: Harmlessness from AI feedback","author":"Bai","year":"2022","journal-title":"arXiv:2212.08073"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.00786"},{"key":"ref19","article-title":"Step-aware preference optimization: Aligning preference with denoising performance at each step","author":"Liang","year":"2024","journal-title":"arXiv:2406.04314"},{"key":"ref20","article-title":"Training diffusion models with reinforcement learning","volume-title":"Proc. 12th Int. Conf. Learn. Represent. (ICLR)","author":"Black"},{"key":"ref21","article-title":"Human preference score V2: A solid benchmark for evaluating human preferences of text-to-image synthesis","author":"Wu","year":"2023","journal-title":"arXiv:2306.09341"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1594"},{"key":"ref23","article-title":"Avoiding mode collapse in diffusion models fine-tuned with reinforcement learning","author":"Barcel\u00f3","year":"2024","journal-title":"arXiv:2410.08315"},{"key":"ref24","article-title":"Tuning-free alignment of diffusion models with direct noise optimization","author":"Tang","year":"2024","journal-title":"arXiv:2405.18881"},{"key":"ref25","article-title":"The unlocking spell on base LLMs: Rethinking alignment via in-context learning","volume-title":"Proc. 12th Int. Conf. Learn. Represent. (ICLR)","author":"Lin"},{"key":"ref26","first-page":"55006","article-title":"LIMA: Less is more for alignment","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"36","author":"Zhou"},{"key":"ref27","article-title":"How robust is GPT-3.5 to predecessors? A comprehensive study on language understanding tasks","author":"Chen","year":"2023","journal-title":"arXiv:2303.00293"},{"key":"ref28","article-title":"Revisiting the superficial alignment hypothesis","author":"Raghavendra","year":"2024","journal-title":"arXiv:2410.03717"},{"key":"ref29","article-title":"Is DPO superior to PPO for LLM alignment? A comprehensive study","author":"Xu","year":"2024","journal-title":"arXiv:2404.10719"},{"key":"ref30","article-title":"SEFE: Superficial and essential forgetting eliminator for multimodal continual instruction tuning","author":"Chen","year":"2025","journal-title":"arXiv:2505.02486"},{"key":"ref31","article-title":"Good seed makes a good crop: Discovering secret seeds in text-to-image diffusion models","author":"Xu","year":"2024","journal-title":"arXiv:2405.14828"},{"key":"ref32","first-page":"6840","article-title":"Denoising diffusion probabilistic models","volume-title":"Proc. NIPS","volume":"33","author":"Ho"},{"key":"ref33","first-page":"1","article-title":"Score-based generative modeling through stochastic differential equations","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Song"},{"key":"ref34","article-title":"Improving image generation with better captions","author":"Goh","year":"2023"},{"key":"ref35","article-title":"Hierarchical text-conditional image generation with CLIP latents","author":"Ramesh","year":"2022","journal-title":"arXiv:2204.06125"},{"key":"ref36","first-page":"8821","article-title":"Zero-shot text-to-image generation","volume-title":"Proc. Int. Conf. Mach. Learn. (ICML)","author":"Ramesh"},{"key":"ref37","first-page":"36479","article-title":"Photorealistic text-to-image diffusion models with deep language understanding","volume-title":"Proc. NIPS","volume":"35","author":"Saharia"},{"key":"ref38","first-page":"8633","article-title":"Video diffusion models","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"35","author":"Ho"},{"key":"ref39","article-title":"High-resolution image synthesis with latent diffusion models","author":"Rombach","year":"2021","journal-title":"arXiv:2112.10752"},{"key":"ref40","first-page":"12606","article-title":"Scaling rectified flow transformers for high-resolution image synthesis","volume-title":"Proc. 41st Int. Conf. Mach. Learn.","author":"Esser"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02161"},{"key":"ref42","article-title":"Margin-aware preference optimization for aligning diffusion models without reference","author":"Hong","year":"2024","journal-title":"arXiv:2406.06424"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.626"},{"key":"ref44","first-page":"37097","article-title":"Diffusion rejection sampling","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Na"},{"key":"ref45","article-title":"Very deep convolutional networks for large-scale image recognition","author":"Simonyan","year":"2014","journal-title":"arXiv:1409.1556"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00371"},{"key":"ref47","article-title":"Openclip","author":"Ilharco","year":"2021"},{"key":"ref48","first-page":"304","article-title":"Linking losses for density ratio and class-probability estimation","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Menon"},{"key":"ref49","first-page":"271","article-title":"F-GAN: Training generative neural samplers using variational divergence minimization","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Nowozin"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1080\/01621459.2000.10473908"},{"key":"ref51","article-title":"SDXL: Improving latent diffusion models for high-resolution image synthesis","volume-title":"Proc. 12th Int. Conf. Learn. Represent. (ICLR)","author":"Podell"},{"key":"ref52","article-title":"Efficient scaling of diffusion transformers for text-to-image generation","author":"Li","year":"2024","journal-title":"arXiv:2412.12391"},{"key":"ref53","first-page":"8748","article-title":"Learning transferable visual models from natural language supervision","volume-title":"Proc. Int. Conf. Mach. Learn.","volume":"139","author":"Radford"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0700"},{"key":"ref55","volume-title":"LAION-Aesthetics","author":"Schuhmann","year":"2022"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2022\/132"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1109\/TMM.2015.2479916"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.emnlp-main.595"},{"key":"ref59","article-title":"CADS: Unleashing the diversity of diffusion models through condition-annealed sampling","volume-title":"Proc. 12th Int. Conf. Learn. Represent. (ICLR)","author":"Sadat"},{"key":"ref60","first-page":"1","article-title":"Particle guidance: Non-I.I.D. diverse sampling with diffusion models","volume-title":"Proc. 12th Int. Conf. Learn. Represent.","author":"Corso"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73027-6_23"},{"key":"ref62","first-page":"30911","article-title":"Shielded diffusion: Generating novel and diverse images using sparse repellency","volume-title":"Proc. 42nd Int. Conf. Mach. Learn.","author":"Kirchhof"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.02170"},{"key":"ref64","article-title":"DPM-Solver++: Fast solver for guided sampling of diffusion probabilistic models","author":"Lu","year":"2022","journal-title":"arXiv:2211.01095"},{"key":"ref65","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1926"},{"key":"ref66","article-title":"GPT-4o system card","author":"Hurst","year":"2024","journal-title":"arXiv:2410.21276"},{"key":"ref67","article-title":"Microsoft COCO captions: Data collection and evaluation server","author":"Chen","year":"2015","journal-title":"arXiv:1504.00325"},{"key":"ref68","article-title":"Flow straight and fast: Learning to generate and transfer data with rectified flow","volume-title":"Proc. 11th Int. Conf. Learn. Represent. (ICLR)","author":"Liu"},{"key":"ref69","first-page":"1","article-title":"Flow matching for generative modeling","volume-title":"Proc. 11th Int. Conf. Learn. Represent.","author":"Lipman"},{"key":"ref70","volume-title":"Flux","author":"Labs","year":"2024"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.1007\/s10463-008-0197-x"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1201\/b14835"},{"key":"ref73","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00387"},{"key":"ref75","first-page":"4596","article-title":"Adafactor: Adaptive learning rates with sublinear memory cost","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Shazeer"},{"key":"ref76","first-page":"1","article-title":"LoRA: Low-rank adaptation of large language models","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Hu"}],"container-title":["IEEE Transactions on Image Processing"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/83\/11355710\/11489281.pdf?arnumber=11489281","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T05:05:05Z","timestamp":1778130305000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11489281\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":76,"URL":"https:\/\/doi.org\/10.1109\/tip.2026.3684401","relation":{},"ISSN":["1057-7149","1941-0042"],"issn-type":[{"value":"1057-7149","type":"print"},{"value":"1941-0042","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026]]}}}