👋 I am an assistant professor at Unprecedented-scale Data Analytics Center, Tohoku University (東北大学 未踏スケールデータアナリティクスセンター). Previously, I received my Ph.D. in Engineering from Tohoku University , M.S. from HoHai University in Jun. 2021, B.S. from JiShou University in Jun. 2018.
🔬 My research interests include 3D vision, scene understanding, and visual perception, with previous work on avatar and image recognition. More recently, I am interested incomputer vision for robotics. I am open to collaborations and can be reached at juan.wang.b1[AT]tohoku.ac.jp (replace [AT] with @).
📢 News
📝 Publications
Journal
Class-agnostic 3D Segmentation by Granularity-Consistent Automatic 2D Mask Tracking.
Juan Wang, Yasutomo Kawanishi, Tomo Miyazaki, Zhijie Wang, Shinichiro Omachi.
-The research advances annotation-free 3D instance segmentation by tracking 2D masks across video and training progressively, turning noisy frame-wise pseudo labels into coherent 3D supervision for accurate, open-vocabulary results.
TAMC: Textual Alignment and Masked Consistency for Open-Vocabulary 3D Scene Understanding.
Juan Wang, Zhijie Wang, Tomo Miyazaki, Yaohou Fan, Shinichiro Omachi.
- The research improves 3D scene understanding through masked consistency training and pseudo-text labels, effectively addressing sparse point cloud processing and training-inference inconsistency issues.
An end-to-end deep neural network is proposed to detect aluminum profile defects \ Wang J, Meng Z H.
- Proposed a Deformable Feature Pyramid module to detect aluminum profile defects.
Conference
ICONIP 2025Scene Text Reconstructor: A Contextual-Aware Masking Framework for Pre-training Text Detectors(Soptlight). Yaohou Fan, Tomo Miyazaki, Zhengmi Tang, Juan Wang, Yongsong Huang, Shinichiro Omachi. 32nd International Conference on Neural Information Processing 35th Annual Meeting of Japanese Neural Network Society.MIRU 2025Class-Agnostic 3D Segmentation without Manual Labels by 2D Mask Tracking. Juan Wang, Yasutomo Kawanishi, Zhijie Wang, Tomo Miyazaki, Shinichiro Omachi. 画像の認識・理解シンポジウム MIRU, 2025.-
MIRU 2024Improved Open-Vocabulary 3D Scene Understanding via Masked Feature Alignment. Juan Wang, Zhijie Wang, Tomo Miyazaki, Shinichiro Omachi. 画像の認識・理解シンポジウム MIRU, 2024.
💻 Work Experience
- Computer Vision Research Intern - Baidu Inc., Global Business Unit
- Content: Utilized diffusion models and deep learning technologies to develop intelligent image/video generation algorithms, delivering AI-powered visual content creation and image optimization solutions for global product offerings.
- Time: Sep. 2025 - Feb.2026.
- Place: Shenzhen, China.
- Research Intern - RGP, RIKEN
- Content: Conducted research on indoor 3D scene understanding under the supervision of Prof. Kawanishi during an internship with the Multimodal Data Recognition Research Team, part of the Guardian Robot Project (GRP) at RIKEN(理化学研究所).
- Time: May. 2025 - Aug.2025.
- Place: Remote-Kyoto, Japan.
- Teaching Assistance - GSIS, Tohoku University
- Course: Data Engineering Data & Science Programming Fundamentals
- Time: Mar. 2024 - Jul.2024.
- Place: Sendai, Japan.
- Software Engineer - Technology Center in 37 Interactive Entertainment
- Content: Developed an automated platform integrating multiple overseas ad channels (TikTok, Google, Line) with real-time performance tracking and alerts; implemented data-driven ad deployment strategies based on ad metrics, like ROI and CPA, while facilitating systematic documentation for future campaign optimization.
- Time: Jul. 2021 - Jan. 2023.
- Place: Guangzhou, China.
🎖 Honors and Awards
- 2023.04 Funding Support from WISE Program for AI Electronics, Tohoku University (Top 1%)
- 2021.06 Outstanding Graduate Student, Hohai University (Top 3%)
- 2018.06 Outstanding Graduate Student, Jishou University (Top 3%)
🧑🎨 Skills
- Python, Pytorch, Java, Php, HTML, CSS, JavaScript, MySQL, Linux.
- Language: English(TOEIC 755, Jan.2023; Duolingo 115, Step.2024), Japanese 日本语(Beginner, 初心者), Chinese (Native).