Luming Tang (唐路明)

I am a Research Scientist at Google DeepMind in New York City. I completed my Ph.D. in Computer Science at Cornell University in 2024, advised by Professor Bharath Hariharan. Prior to that, I earned my Bachelor degree in Mathematics and Physics from Tsinghua University in China.

Email / GitHub / Google Scholar / Twitter / LinkedIn

Research

(* indicates equal contribution)

	Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities Gemini Team Tech Report, 2025 arxiv
	Unified Autoregressive Visual Generation and Understanding with Continuous Tokens Lijie Fan, Luming Tang, Siyang Qin, Tianhong Li, Xuan Yang, Siyuan Qiao, Andreas Steiner, Chen Sun, Yuanzhen Li, Tao Zhu, Michael Rubinstein, Michalis Raptis, Deqing Sun, Radu Soricut Tech Report*, 2025 arxiv
	Coarse Correspondences Boost Spatial-Temporal Reasoning in Multimodal Language Model Benlin Liu, Yuhao Dong, Yiqin Wang, Zixian Ma, Yansong Tang, Luming Tang, Yongming Rao, Wei-Chiu Ma, Ranjay Krishna CVPR, 2025 paper
	RealFill: Reference-Driven Generation for Authentic Image Completion Luming Tang, Nataniel Ruiz, Qinghao Chu, Yuanzhen Li, Aleksander Holynski, David E. Jacobs, Bharath Hariharan, Yael Pritch, Neal Wadhwa, Kfir Aberman, Michael Rubinstein SIGGRAPH, 2024 (Journal Track) paper / video / project page
	Emergent Correspondence from Image Diffusion Luming Tang, Menglin Jia, Qianqian Wang, Cheng Perng Phoo, Bharath Hariharan NeurIPS*, 2023 paper / video / code / poster / slides / project page
	Magic3D: High-Resolution Text-to-3D Content Creation Chen-Hsuan Lin, Jun Gao, Luming Tang, Towaki Takikawa, Xiaohui Zeng, Xun Huang, Karsten Kreis, Sanja Fidler, Ming-Yu Liu, Tsung-Yi Lin CVPR, 2023 (Highlight)* paper / project page
	Visual Prompt Tuning Menglin Jia, Luming Tang, Bor-Chun Chen, Claire Cardie, Serge Belongie, Bharath Hariharan, Ser-Nam Lim ECCV, 2022 paper / video / code
	Few-Shot Classification with Feature Map Reconstruction Networks Davis Wertheimer, Luming Tang, Bharath Hariharan CVPR, 2021 paper / video / code / poster
	Revisiting Pose-Normalization for Fine-Grained Few-Shot Recognition Luming Tang, Davis Wertheimer, Bharath Hariharan CVPR, 2020 paper / video / code / slides
	Multi-Entity Dependence Learning with Rich Context via Conditional Variational Auto-encoder Luming Tang, Yexiang Xue, Di Chen, Carla P. Gomes AAAI, 2018 paper / code / poster / slides
	Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification Zhongdao Wang, Luming Tang, Xihui Liu, Zhuliang Yao, Shuai Yi, Jing Shao, Junjie Yan, Shengjin Wang, Hongsheng Li, Xiaogang Wang ICCV, 2017 paper / dataset / poster
	Hierarchical Deep Recurrent Architecture for Video Understanding Luming Tang, Boyang Deng, Haiyu Zhao, Shuai Yi CVPR Workshop on Youtube-8M Large-Scale Video Understanding, 2017 paper / code

Projects

Here're some interesting research or course projects I have worked on.

	Diagnosing and Remedying Shot Sensitivity with Cosine Few-Shot Learners Davis Wertheimer, Luming Tang, Bharath Hariharan Tech Report, 2022 arxiv
	Few-Shot Learning in Long-Tailed Settings Davis Wertheimer, Luming Tang, Dhruv Baijal, Pranjal Mittal, Anika Talwar, Bharath Hariharan Tech Report, 2021 paper / code
	Baseline implementation of DrQA and BERT finetuning on SQuAD 2.0 Luming Tang CS 5740 Natural Language Processing, Assignment 4, 2020 code
	An Experimental Evaluation of Optimized Maximum Flow Implementations Junxi Song, Luming Tang, Hongbo Zhang, Xiaoji Zhang (alphabetical order) CS 6820 Analysis of Algorithms, Course Project, 2019 draft
	Garbage Collection Schedule Algorithm Luming Tang, Junxi Song CS 6820 Analysis of Algorithms, Assignment 2 Problem 4, 2019 problem / our proof
	On the Regularization Balance in Autoencoder-Based Generative Models Bin Dai, Luming Tang, David Wipf Tech Report, 2019 draft / supplementary
	OpenNRE: An Open-Source Package for Neural Relation Extraction Tianyu Gao, Xu Han, Shulin Cao, Luming Tang, Yankai Lin, Zhiyuan Liu THUNLP Github, 2018 code

Service

Teaching Assistant:

Spring 2019, CS 4787 Principles of Large-Scale Machine Learning
Summer 2019, CS 2110 OOP and Data Structures
Fall 2019, CS 6670 Graduate Computer Vision

PhD admission committee student volunteer: 2020, 2023

Conference Reviewer: CVPR, ICCV, ECCV, ICML, NeurIPS, ICLR, SIGGRAPH Asia, SIGGRAPH

Journal Reviewer: TPAMI, IJCV

PC member: AAAI'23

Misc.

I love playing soccer and FIFA (FIFA'20 Season Division 1 with Title, FIFA'22 Ultimate Team Division 3). Unfortunately, right now I'm fighting against a bad ankle injury. Wish I could come back to soccer soon!
Meet the most adorable S'more and check out her instagram! Her Chinese name is 屎妹 :)
I am from Jiaozuo, a beautiful small city in Henan, China.

Design and source code modified based on Leonid Keselman and Jon Barron's website

last update: Aug, 2025

Luming Tang (唐路明)

Research

Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities

Unified Autoregressive Visual Generation and Understanding with Continuous Tokens

Coarse Correspondences Boost Spatial-Temporal Reasoning in Multimodal Language Model

RealFill: Reference-Driven Generation for Authentic Image Completion

Emergent Correspondence from Image Diffusion

Magic3D: High-Resolution Text-to-3D Content Creation

Visual Prompt Tuning

Few-Shot Classification with Feature Map Reconstruction Networks

Revisiting Pose-Normalization for Fine-Grained Few-Shot Recognition

Multi-Entity Dependence Learning with Rich Context via Conditional Variational Auto-encoder

Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification

Hierarchical Deep Recurrent Architecture for Video Understanding

Projects

Diagnosing and Remedying Shot Sensitivity with Cosine Few-Shot Learners

Few-Shot Learning in Long-Tailed Settings

Baseline implementation of DrQA and BERT finetuning on SQuAD 2.0

An Experimental Evaluation of Optimized Maximum Flow Implementations

Garbage Collection Schedule Algorithm

On the Regularization Balance in Autoencoder-Based Generative Models

OpenNRE: An Open-Source Package for Neural Relation Extraction

Service

Misc.