图像生成

HADIS: Hybrid Adaptive Diffusion Model Serving for Efficient Text-to-Image Generation

HADIS：混合自适应扩散模型服务用于高效的文本到图像生成

Qizheng Yang, Tung-I Chen, Siyu Zhao, Ramesh K. Sitaraman, Hui Guan

专题命中 文生图 ：text-to-image(title,abstract)；diffusion(title,abstract)；image generation(title)

AI总结 HADIS通过混合自适应架构优化扩散模型服务，提升响应质量和降低延迟违规率

Comments 15 pages, 12 figures

How Many Images Does It Take? Estimating Imitation Thresholds in Text-to-Image Models

需要多少张图片？文本到图像模型中模仿阈值的估计

Sahil Verma, Royi Rassin, Arnav Das, Gantavya Bhatt, Preethi Seshadri, Chirag Shah, Jeff Bilmes, Hannaneh Hajishirzi, Yanai Elazar

机构 * University of Washington, Seattle（华盛顿大学） ； Bar-Ilan University（巴伊兰大学） ； University of California, Irvine（加州大学伊文斯顿分校） ； Allen Institute of AI（人工智能研究院） 纠错

专题命中 文生图 ：text-to-image(title,abstract)；分类 cs.CV

AI总结 研究通过评估文本到图像模型的模仿阈值，探讨其在版权和隐私合规中的应用。

Comments Accepted at TMLR 2025, ATTRIB, RegML, and SafeGenAI workshops at NeurIPS 2024 and NLLP Workshop 2024. https://openreview.net/forum?id=x0qJo7SPhs

AbductiveMLLM: Boosting Visual Abductive Reasoning Within MLLMs

AbductiveMLLM: 提升多模态大语言模型中的视觉归纳推理

Boyu Chang, Qi Wang, Xi Guo, Zhixiong Nan, Yazhou Yao, Tianfei Zhou

专题命中 文生图 ：text-to-image(abstract)；diffusion(abstract)；分类 cs.CV

AI总结 AbductiveMLLM通过结合REASONER和IMAGINER组件，提升多模态大语言模型在视觉归纳推理中的性能。

Comments Accepted by AAAI 2026 as Oral. Code:https://github.com/ChangPtR/AbdMLLM

Topological Perspectives on Optimal Multimodal Embedding Spaces

拓扑视角下的最优多模态嵌入空间

Abdul Aziz A. B, A. B Abdul Rahim

专题命中 文生图 ：image generation(abstract)；text-to-image(abstract)

AI总结 本文通过拓扑数据分析比较CLIP和CLOOB的嵌入空间，揭示其模态差距驱动因素和维度坍缩的影响，为多模态模型优化提供新视角。

Comments This manuscript contains substantive technical inaccuracies and an incomplete treatment of the stated topic. Subsequent developments and a reassessment of the problem indicate that the scope and framing of the work do not adequately reflect the current state of research, and the analysis is therefore incomplete and outdated