Hello 👋, I'm Gabriel (Jiahui) Huang

I’m a Research Scientist at Adobe Research, working on topics related to video processing and editing.

I live in the beautiful city of Vancouver, Canada 🇨🇦, where I previously received my Master’s degree from UBC under Leonid Sigal.

Scholar LinkedIn Email

Highlights

Jan 2026: Shipped “Shape Mask Tracker” to Adobe Premiere Pro .
Oct 2024: Presented project “Clean Machine” at Adobe Max Sneaks 2024 .
Oct 2023: Presented project “Fast Fill” at Adobe Max Sneaks 2023 .
Oct 2022: Internship project “Instant Add” show-cased at Adobe Max Sneaks 2022 .

Publications

DAGE: Dual-Stream Architecture for Efficient and Fine-Grained Geometry Estimation

DAGE: Dual-Stream Architecture for Efficient and Fine-Grained Geometry Estimation

Tuan Duc Ngo, Jiahui Huang, Seoung Wug Oh, Kevin Blackburn-Matzen, Evangelos Kalogerakis, Chuang Gan, Joon Young Lee

CVPR, 2026

A dual-stream transformer that decouples global view consistency from fine-detail preservation for high-resolution multi-view geometry estimation.

Project Page Paper Code

Generative Video Motion Editing with 3D Point Tracks

Generative Video Motion Editing with 3D Point Tracks

Yao-Chih Lee, Zhoutong Zhang, Jiahui Huang, Jui-Hsien Wang, Joon Young Lee, Jia-Bin Huang, Eli Shechtman, Zhengqi Li

CVPR, 2026

A framework for precise video motion editing via 3D point tracks, enabling joint control over camera and object motion.

Project Page Paper

VideoMaMa: Mask-Guided Video Matting via Generative Prior

VideoMaMa: Mask-Guided Video Matting via Generative Prior

Sangbeom Lim, Seoung Wug Oh, Jiahui Huang, Heeji Yoon, Seungryong Kim, Joon Young Lee

CVPR, 2026

Turns coarse segmentation masks into pixel-accurate alpha mattes by leveraging pretrained video diffusion priors, with zero-shot generalization to in-the-wild footage.

Project Page Paper Code

AnthroTAP: Learning Point Tracking with Real-World Motion

AnthroTAP: Learning Point Tracking with Real-World Motion

Inès Hyeonsu Kim, Seokju Cho, Jahyeok Koo, Junghyun Park, Jiahui Huang, Honglak Lee, Joon Young Lee, Seungryong Kim

CVPR, 2026

Generates pseudo-labeled point-tracking training data by exploiting human-motion complexity captured via the SMPL model.

Project Page Paper Code

Seurat: From Moving Points to Depth

Seurat: From Moving Points to Depth

Seokju Cho, Jiahui Huang, Seungryong Kim, Joon Young Lee

CVPR, 2025 Highlight

Inferring depth from point trajectories

Project Page Paper Code

Exploring Temporally-Aware Features for Point Tracking

Exploring Temporally-Aware Features for Point Tracking

Inès Hyeonsu Kim, Seokju Cho, Jiahui Huang, Jung Yi, Joon Young Lee, Seungryong Kim

CVPR, 2025

A feature backbone for point tracking with built-in temporal awareness

Project Page Paper Code

MotionCanvas: Cinematic Shot Design with Controllable Image-to-Video Generation

MotionCanvas: Cinematic Shot Design with Controllable Image-to-Video Generation

Jinbo Xing, Long Mai, Cusuh Ham, Jiahui Huang, Aniruddha Mahapatra, Chi-Wing Fu, Tien-Tsin Wong, Feng Liu

SIGGRAPH, 2025

A model that integrates user-driven controls into image-to-video (I2V) generation models

Project Page Paper

LocoTrack: Local All-Pair Correspondence for Point Tracking

LocoTrack: Local All-Pair Correspondence for Point Tracking

Seokju Cho, Jiahui Huang, Jisu Nam, Honggyu An, Seungryong Kim, Joon Young Lee

ECCV, 2024

A highly accurate and efficient model for point tracking.

Project Page Paper Code

FlowTrack: Revisiting Optical Flow for Long-Range Dense Tracking

FlowTrack: Revisiting Optical Flow for Long-Range Dense Tracking

Seokju Cho, Jiahui Huang, Seungryong Kim, Joon Young Lee

CVPR, 2024

Dense point tracking on long-range videos.

Paper

INVE: Interactive Neural Video Editing

INVE: Interactive Neural Video Editing

Jiahui Huang, Leonid Sigal, Kwang Moo Yi, Oliver Wang, Joon Young Lee

CGI, 2023

A real-time video editing pipeline via edits propagation.

Project Page Paper

Layered Controllable Video Generation

Layered Controllable Video Generation

Jiahui Huang, Yuhe Jin, Kwang Moo Yi, Leonid Sigal

ECCV, 2022 Oral

Using masks to control the generation of videos.

Project Page Paper Code

Single Image Video Prediction with Auto-Regressive GANs

Jiahui Huang, Yew Ken Chia, Samson Yu, Kevin Yee, Dennis Küster, Eva G Krumhuber, Dorien Herremans, Gemma Roig

Sensors, Volume 22

Animating facial expressions using GANs in an auto-regressive fashion.

Duality diagram similarity: a generic framework for initialization selection in task transfer learning

Kshitij Dwivedi, Jiahui Huang, Gemma Roig

ECCV, 2020

A metric that measures the similarity between neural networks.

Deep Anchored Convolutional Neural Networks

Jiahui Huang, Kshitij Dwivedi, Gemma Roig

CVPR Workshops, 2019 Oral

A highly efficitn neural network architecture that reuses weights from different layres.