Yuqi Wang 王宇奇

I build controllable generative systems for images and video.

Editorial still life of film strips, optical glass, geometric studies, and a cobalt audio waveform

Publications

Unicorn architecture with shot-aware planning, in-context keyframe editing, and keyframe-anchored video editing
AAAI 2027 / Under review First Author

Unicorn: Controllable Multi-Shot Video Editing Framework via Shot-wise Keyframe Anchors

Yuqi Wang et al.

SREdit architecture with semantic instruction atomization, aligned self-reward, and spatial reward reweighting
ACM MM 2026 / Poster First Author

SREdit: A Self-Rewarding Framework with Intrinsically Aligned Critic for Image Editing

Yuqi Wang et al.

Paper
FedPDG architecture with personalized diffusion generation, prediction discrepancy selection, and client data augmentation
ICML 2026 / Poster First Author

FedPDG: Prediction Discrepancy-Guided Data Generation for Heterogeneous Federated Learning

Yuqi Wang et al.

Paper

Experience

Beijing

Kuaishou 快手

AIGC Algorithm Intern / Commercial Technology, Safety & Ecosystem

Researching controllable multi-shot video editing and building MultiShotBench.

Beijing

Meituan 美团

AIGC Algorithm Intern / Digital Human & Creative Generation

Built image composition and automatic video editing systems serving 1,000+ videos daily.

I study how models see, edit, and preserve.

I’m a master’s student in Software Engineering at Beihang University, advised by Prof. Jianwei Niu in the LDMC Lab. I work on diffusion, multimodal models, and reinforcement learning for visual generation.

Education

M.S. in Software Engineering

Beihang University / LDMC Lab

B.E. in Software Engineering

Beihang University / Outstanding Graduate

Interests

Yuqi Wang brand artwork combining circular research diagrams and an audio waveform

R&B and hip-hop

Music, production, and sampling.

Philosophy

Mind. Freedom. Meaning. Action.

应无所住,
而生其心

生活

随手记些日常。

看看
Circular research forms and a cobalt audio waveform

生活区,开张。

随笔

留白,也是一种开始。

阅读

Say hello.