Lu Ling

I am passionate about multimodal reasoning and self-improving AI agents. I lead DL3DV-10K (340k ↓), a large-scale scene dataset. I build controllable world models that enable agents to generate, understand, and reason about physical environments, with the goal of advancing virtual content creation and embodied intelligence.

I earned my Ph.D. from Purdue University. I am a Member of Technical Staff at World Labs and was a research intern at NVIDIA Cosmos Lab in summer and fall 2024, where I had the good fortune to work with Max (Zhaoshuo) Li, Chen-Hsuan Lin, Tsung-Yi Lin, and Ming-Yu Liu.

Email / Google Scholar / LinkedIn / Twitter

News

[09/2026] OptiWorld: Optimal Control for Video World Generation under Physical Constraints, has been accepted by NeurIPS 2026!
[02/2026] I-Scene: an interactive 3D scene foundation model, has been accepted by CVPR 2026!
[01/2026] Scenethesis: an agentic framework for 3D scene generation, has been accepted by ICLR 2026!
[06/2025] DL3DV-10K is selected as CVPR 2025 demo presentation.
[06/2025] Invited talk at World Labs.
[04/2024] DL3DV-10K has been presented at CVPR 2024 workshops including AI4CC, Urban Scene Modeling USM3D, and Multimodalities for 3D Scenes.
[02/2024] DL3DV-10K has been accepted by CVPR 2024!
[06/2023] Dr.Bokeh and PixHt-Lab have been accepted by CVPR 2023!

Selected Publications

	I-Scene: 3D Instance Models are Implicit Generalizable Spatial Learners Lu Ling, Yunhao Ge, Yichen Sheng, Aniket Bera CVPR 2026 Feed-forward interactive 3D scene generation / world models. Trained on a small amount of data while outperforming SAM3D in generalizability to unseen layouts and spatial relations. > proj > paper > code
	Scenethesis: A Language and Vision Agentic Framework for 3D Scene Generation Lu Ling, Chen-Hsuan Lin, Tsung-Yi Lin, Yifan Ding, Yu Zeng, Yichen Sheng, Yunhao Ge, Ming-Yu Liu, Aniket Bera, Zhaoshuo Li ICLR 2026 An Agentic Framework for Text-to-3D Scene Generation. The generated scenes are diverse, interactive, realistic, and physically plausible for virtual content creation, editing, simulation, and embodied AI. > proj > paper
	DL3DV-10K: A Large-Scale Scene Dataset for Deep Learning-based 3D Vision Lu Ling, Yichen Sheng, Zhi Tu, Wentian Zhao, Cheng Xin, Kun Wan, Lantao Yu, Qianyu Guo, Zixun Yu, Yawen Lu, Xuanmao Li, Xingpeng Sun, Rohan Ashok, Aniruddha Mukherjee, Hao Kang, Xiangrui Kong, Gang Hua, Tianyi Zhang, Bedrich Benes, Aniket Bera CVPR 2024 > proj > paper > code
	Dr.Bokeh: DiffeRentiable Occlusion-aware Bokeh Rendering Yichen Sheng, Zixun, Yu, Lu Ling, Zhiwen Cao, Cecilia Zhang, Xin Lu, Ke Xian, Haiting Lin, Bedrich Benes CVPR 2024 Lens blur effects in image compositing > proj > video
	PixHt-Lab: Pixel Height Based Light Effect Generation for Image Compositing Yichen Sheng, Jianming Zhang, Julien Philip, Yannick Hold-Geoffroy, Xin Sun, HE Zhang, Lu Ling, Bedrich Benes CVPR 2023 (highlight) Light effects synthesis in image compositing > paper > code
	Cooperating Graph Neural Networks with Deep Reinforcement Learning for Vaccine Prioritization Lu Ling, Washim Uddin Mondal, Satish V. Ukkusuri IEEE Journal of Biomedical and Health Informatics (2024) Reinforcement learning, graph neural network, deep learning, disease diffusion > paper