Bingchuan Li
Portrait
Logo Senior Researcher at ByteDance
Logo Master in Bupt

I am a Senior Researcher in the Intelligent Creation department at ByteDance, where I have been working since graduating with my Master's degree from Beijing University of Posts and Telecommunications (BUPT) in 2020.

My professional journey is deeply rooted in the fascinating field of Generative AI. During the era of Generative Adversarial Networks (GANs), I embarked on pioneering research in visual synthesis and manipulation. As the emergence of Diffusion models has driven the evolution of the technological landscape, I have consistently remained at the forefront of exploring controllable visual generation, striving to achieve unparalleled realism and precision. Currently, I am dedicated to tackling the core challenges in the dynamic domain of AI video generation — a pursuit I firmly believe is a crucial pathway toward the grand vision of Artificial General Intelligence (AGI).

Please feel free to reach out for collaborations, internships, or full-time opportunities.

Google Scholar GitHub WeChat LinkedIn
Education
  • Beijing University of Posts and Telecommunications
    Beijing University of Posts and Telecommunications
    M.S. in Information and Communication Engineering
    Sep. 2017 - Jul. 2020
  • Tiangong University
    Tiangong University
    B.S. in Electronic Information
    Sep. 2013 - Jul. 2017
Experience
  • ByteDance
    ByteDance
    Multimodal Generation Researcher
    Jun. 2020 - Present
  • Horizon Robotics
    Horizon Robotics
    Research Intern
    Nov. 2019 - Mar. 2020
  • SenseTime
    SenseTime
    Research Intern
    Mar. 2019 - Aug. 2019
News
2025
We release the Phantom dataset, built upon koala-36M, on Huggingface Phantom-data-Koala36M. Stars
Oct 13
We release the Project Page and Technique Report of OmniInsert.
Sep 24
We release the project page, technical report, model weights, and inference code for HuMo. Stars
Sep 11
We are delighted that Phantom, OneGT has been accepted to the ICCV 2025. 🎉🎉🎉
Jul 16
We are delighted that HyperLoRA has been accepted to the CVPR 2025. 🎉🎉🎉
Feb 27
We proposed a novel subject-consistent video generation model, Phantom, and have released the report publicly. Stars
Feb 17
2023
Congratulations! GaFET has been accepted to ICCV 2023! 🎉🎉
Jul 24
Three papers accepted to AAAI 2023! 🎉🎉🎉 Oral
Feb 08
Selected Publications (view all )
(* equal contribution, project lead)
HuMo: Human-Centric Video Generation via Collaborative Multi-Modal Conditioning
HuMo: Human-Centric Video Generation via Collaborative Multi-Modal Conditioning

Liyang Chen*, Tiangxiang Ma*, Jiawei Liu, Bingchuan Li, Zhuowei Chen, Lijie Liu, Xu He, Gen Li, Qian He, Zhiyong Wu

AAAI 2026

We introduce HuMo, a unified Human-Centric Video Generation framework that overcomes multimodal coordination challenges through a new high-quality dataset and a progressive training paradigm, achieving state-of-the-art subject preservation and audio-visual sync.

[Paper] [Code] [Demo]

HuMo: Human-Centric Video Generation via Collaborative Multi-Modal Conditioning

Liyang Chen*, Tiangxiang Ma*, Jiawei Liu, Bingchuan Li, Zhuowei Chen, Lijie Liu, Xu He, Gen Li, Qian He, Zhiyong Wu

AAAI 2026

We introduce HuMo, a unified Human-Centric Video Generation framework that overcomes multimodal coordination challenges through a new high-quality dataset and a progressive training paradigm, achieving state-of-the-art subject preservation and audio-visual sync.

[Paper] [Code] [Demo]

Phantom-Data: Towards a General Subject-Consistent Video Generation Dataset
Phantom-Data: Towards a General Subject-Consistent Video Generation Dataset

Zhuowei Chen*, Bingchuan Li*, Tiangxiang Ma*, Lijie Liu*, Mingcong Liu, Yi Zhang, Gen Li, Xinghui Li, Siyu Zhou, Qian He, Xinglong Wu

Arxiv 2025

To address the subject-background entanglement in subject-to-video generation, we introduce Phantom-Data, the first general-purpose cross-pair consistency dataset, which significantly improves prompt alignment and visual quality while preserving identity.

[Paper] [Code] [Demo]

Phantom-Data: Towards a General Subject-Consistent Video Generation Dataset

Zhuowei Chen*, Bingchuan Li*, Tiangxiang Ma*, Lijie Liu*, Mingcong Liu, Yi Zhang, Gen Li, Xinghui Li, Siyu Zhou, Qian He, Xinglong Wu

Arxiv 2025

To address the subject-background entanglement in subject-to-video generation, we introduce Phantom-Data, the first general-purpose cross-pair consistency dataset, which significantly improves prompt alignment and visual quality while preserving identity.

[Paper] [Code] [Demo]

OneGT: One-Shot Geometry-Texture Neural Rendering for Head Avatars
OneGT: One-Shot Geometry-Texture Neural Rendering for Head Avatars

Jinshu Chen, Bingchuan Li, Feiwei Zhang, Songtao Zhao, Qian He

ICCV 2025

We present OneGT that adheres to the frameworks of the rendering tools, while restructuring individual stages of the rendering pipeline through neural networks

[paper]

OneGT: One-Shot Geometry-Texture Neural Rendering for Head Avatars

Jinshu Chen, Bingchuan Li, Feiwei Zhang, Songtao Zhao, Qian He

ICCV 2025

We present OneGT that adheres to the frameworks of the rendering tools, while restructuring individual stages of the rendering pipeline through neural networks

[paper]

Phantom: Subject-Consistent Video Generation via Cross-Modal Alignment
Phantom: Subject-Consistent Video Generation via Cross-Modal Alignment

Lijie Liu*, Tiangxiang Ma*, Bingchuan Li*, Zhuowei Chen*, Jiawei Liu, Gen Li, Siyu Zhou, Qian He, Xinglong Wu

ICCV 2025 Spotlight

A unified framework that learns cross-modal alignment from text-image-video triplets to achieve high-fidelity, subject-consistent videos while resolving content leakage and multi-subject confusion.

[Paper] [Code] [Demo]

Phantom: Subject-Consistent Video Generation via Cross-Modal Alignment

Lijie Liu*, Tiangxiang Ma*, Bingchuan Li*, Zhuowei Chen*, Jiawei Liu, Gen Li, Siyu Zhou, Qian He, Xinglong Wu

ICCV 2025 Spotlight

A unified framework that learns cross-modal alignment from text-image-video triplets to achieve high-fidelity, subject-consistent videos while resolving content leakage and multi-subject confusion.

[Paper] [Code] [Demo]

HyperLoRA: Parameter-Efficient Adaptive Generation for Portrait Synthesis
HyperLoRA: Parameter-Efficient Adaptive Generation for Portrait Synthesis

Mengtian Li, Jinshu Chen, Wanquan Feng, Bingchuan Li, Fei Dai, Songtao Zhao, Qian He

CVPR 2025 Spotlight

We introduce HyperLoRA, a parameter-efficient method that generates adaptive LoRA weights to achieve high-fidelity, zero-shot personalized portrait synthesis, merging the high performance of LoRA with the zero-shot capability of adapter-based techniques.

[Code] [Paper] [Demo]

HyperLoRA: Parameter-Efficient Adaptive Generation for Portrait Synthesis

Mengtian Li, Jinshu Chen, Wanquan Feng, Bingchuan Li, Fei Dai, Songtao Zhao, Qian He

CVPR 2025 Spotlight

We introduce HyperLoRA, a parameter-efficient method that generates adaptive LoRA weights to achieve high-fidelity, zero-shot personalized portrait synthesis, merging the high performance of LoRA with the zero-shot capability of adapter-based techniques.

[Code] [Paper] [Demo]

ReGANIE: Rectifying GAN Inversion Errors for Accurate Real Image Editing
ReGANIE: Rectifying GAN Inversion Errors for Accurate Real Image Editing

Bingchuan Li, Tiangxiang Ma, Peng Zhang, Miao Hua, Wei Liu, Qian He, Zili Yi

AAAI 2023 Oral

To overcome the reconstruction-editability trade-off in StyleGAN inversion, we propose a two-phase framework that first uses an inversion network for editing and then a rectifying network to correct errors, enabling accurate real image manipulation with near-perfect reconstruction.

[paper]

ReGANIE: Rectifying GAN Inversion Errors for Accurate Real Image Editing

Bingchuan Li, Tiangxiang Ma, Peng Zhang, Miao Hua, Wei Liu, Qian He, Zili Yi

AAAI 2023 Oral

To overcome the reconstruction-editability trade-off in StyleGAN inversion, we propose a two-phase framework that first uses an inversion network for editing and then a rectifying network to correct errors, enabling accurate real image manipulation with near-perfect reconstruction.

[paper]

Dystyle: Dynamic neural network for multi-attribute-conditioned style editing
Dystyle: Dynamic neural network for multi-attribute-conditioned style editing

Bingchuan Li*, Shaofei Cai*, Wei Liu, Peng Zhang, Miao Hua, Qian He, Zili Yi

WACV 2023

We design a Dynamic Style Manipulation Network (DyStyle) whose structure and parameters vary by input samples, to perform nonlinear and adaptive manipulation of latent codes for flexible and precise attribute control

[code] [paper]

Dystyle: Dynamic neural network for multi-attribute-conditioned style editing

Bingchuan Li*, Shaofei Cai*, Wei Liu, Peng Zhang, Miao Hua, Qian He, Zili Yi

WACV 2023

We design a Dynamic Style Manipulation Network (DyStyle) whose structure and parameters vary by input samples, to perform nonlinear and adaptive manipulation of latent codes for flexible and precise attribute control

[code] [paper]

All publications