Lune

ICML2025顶会

UniMC: Taming Diffusion Transformer for Unified Keypoint-Guided Multi-Class Image Generation

Qin Guo, Ailing Zeng, Dongxu Yue, Ceyuan Yang, Yang Cao, Hanzhong Guo, Fei Shen, Wei Liu, Xihui Liu, Dan Xu

摘要

UNIMC: Taming Diffusion Transformer for Unified Keypoint-Guided Multi-Class Image Generation ness of UNIMC, particularly in heavy occlusions and multi-class scenarios. Project page can be found at this link. (A) Raw image (B) Most condition form (C) Our condition form Class name: … Keypoint: … Bounding box: … Keypoint binding confusion Class binding confusion (A) Raw image (B) Most condition form (C) Our condition form Keypoint binding confusion Class binding confusion ☹ 🤗

问 Lune

尽管问,回答会引用原文。

Lune 会读完全文和相关的 Paper,每个回答都附上所依据的原文段落。

可以从这些问题问起

Lune 的依据

全文 · 引用它的 3 篇顶会 Paper · 相关的 Paper

获取答案

免费开始,无需绑卡

引用它的顶会 Paper3

  1. M4V: Multimodal Mamba for Efficient Text-to-Video GenerationJiancheng Huang, Gengwei Zhang, Zequn Jie, Siyu Jiao 等CVPR 2026 · 14 次引用
  2. WildActor: Unconstrained Identity-Preserving Video GenerationQin Guo, Tianyu Yang, Xuanhua He, Fei Shen 等ICML 2026 · 4 次引用
  3. OmniPortrait: Fine-Grained Personalized Portrait Synthesis via Pivotal OptimizationDongxu Yue, Bo Lin, Yao Tang, Jiajun Liang 等ICLR 2026

它引用的 Paper31

  1. Learning Transferable Visual Models From Natural Language SupervisionAlec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh 等ICML 2021 · 47,906 次引用
  2. Denoising Diffusion Probabilistic ModelsJonathan Ho, Ajay Jain, Pieter AbbeelNeurIPS 2020 · 35,902 次引用
  3. Segment AnythingAlexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao 等ICCV 2023 · 13,211 次引用
  4. Diffusion Models Beat GANs on Image SynthesisPrafulla Dhariwal, Alexander Quinn NicholNeurIPS 2021 · 13,211 次引用
  5. High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser 等CVPR 2022 · 13,123 次引用

相关 Paper

  1. UNIC-Adapter: Unified Image-instruction Adapter with Multi-modal Transformer for Image GenerationLunhao Duan, Shanshan Zhao, Wenjun Yan, Yinglun Li 等CVPR 2025
  2. Inpaint-Anywhere: Zero-Shot Multi-Identity Inpainting with Efficient Diffusion TransformerJunsheng Luan, Lei Zhao, Wei XingAAAI 2026
  3. One Transformer Fits All Distributions in Multi-Modal Diffusion at ScaleFan Bao, Shen Nie, Kaiwen Xue, Chongxuan Li 等ICML 2023 · 236 次引用
  4. Unicombine: Unified Multi-Conditional Combination with Diffusion TransformerHaoxuan Wang, Jinlong Peng, Qingdong He, Hao Yang 等ICCV 2025 · 6 次引用
  5. Conditional Latent Diffusion Models for Zero-Shot Instance SegmentationMaximilian Ulmer, Wout Boerdijk, Rudolph Triebel, Maximilian DurnerICCV 2025 · 1 次引用
黄昏的海面,两侧是细线勾勒的悬崖