Research author
Xinyang Li
2 AI research papers in the One9Founders library, with summaries and links to original sources.
Papers by Xinyang Li
Aug 14, 2026
cs.AI
Wrong but Useful: Trajectory Value Beyond Answer Correctness in Multi-Agent Messages
Chih-Hsuan Yang, Anjir Ahmed Chowdhury, Cheng-Hau Yang, et al.
Mar 19, 2026
cs.CV
CustomTex: High-fidelity Indoor Scene Texturing via Multi-Reference Customization
Weilin Chen, Jiahao Rao, Wenhao Wang, et al.
CustomTex is a new AI system that creates high-quality textures for 3D indoor scenes by using reference images as examples rather than text descriptions. It works by taking an untextured 3D room and reference images showing how each object should look, then automatically generates detailed texture maps with sharp, artifact-free results. The method uses a dual-distillation technique to balance both semantic accuracy (matching the reference style) and pixel-level quality (making everything look crisp and realistic).
3D generationtexture synthesisscene editingimage-to-3D