Explore · 每周精选
发现最受关注的研究论文,追踪研究趋势,订阅感兴趣的期刊与关键词。
36 篇论文 · 按点赞排序
Zhicheng Sun, Zhenhao Yang, Yang Jin +9 authors
A classifier guidance method resolves limitations in diffusion models for identity-preserving image generation using off-the-shelf image discriminators, enabling flexible personalization across various subjects.
Manoj Kumar, Neil Houlsby, Emiel Hoogeboom
Semantica, an image-conditioned diffusion model, generates high-quality images based on semantic information and adapts to different datasets without fine-tuning.
Dor Verbin, Pratul P. Srinivasan, Peter Hedman +4 authors
The proposed method enhances NeRF's rendering of specular objects by using ray tracing and a small network to decode feature vectors, achieving photorealistic reflections with efficient optimization.
Liwen Wu, Sai Bi, Zexiang Xu +5 authors
Neural Directional Encoding enhances neural radiance fields for rendering specular objects by incorporating angular encoding and spatially varying directional encoding to model reflections and improve real-time performance.
Hanzhang Tu, Ruizhi Shao, Xue Dong +9 authors
Tele-Aloha, a low-budget telepresence system, uses a novel view synthesis algorithm with cascaded disparity estimation and Gaussian Splatting for high-quality, real-time 3D communication without head-mounted displays.
Yudong Li, Yuhao Feng, Wen Zhou +4 authors
ChatFlow, a cross-language transfer-based LLM using Chinese, English, and parallel corpus, enhances training efficiency and performance for Chinese language models leveraging LLaMA2.
北京市昌平区探索星信息技术及软件开发工作室
京ICP备2026059466号