Greetings from Evanston.
I am currently a fourth-year Ph.D. student at Northwestern University under the supervision of Prof. Ying Wu. I am interested in computer vision, multi-modality, embodied AI, and vision-language models. My current research areas are multi-modality and embodied AI, especially active visual recognition with vision-language understanding, which integrates intelligent control strategies into the visual recognition process to solve various recognition challenges. I am constantly investigating the challenges inherent to active vision agents in an open-world context. These challenges include, but are not limited to, multi-modality, vision-language models, and few-sample learning.
In addition to embodied AI, I also have project experience in other real-world vision challenges, including temporal action localization, anomaly detection, and medical image analysis.
My detailed resume/CV is here (last updated on September 2024).
🔥 News
- 2024.05: 🎉🎉 Our paper Outlier-Probability-Based Feature Adaptation for Robust Unsupervised Anomaly Detection on Contaminated Training Data has been accepted by IEEE Transactions on Circuits and Systems for Video Technology!
- 2024.02: 🎉🎉 Our paper Active Open-Vocabulary Recognition: Let Intelligent Moving Mitigate CLIP Limitations has been accepted by CVPR 2024!
📖 Educations
- 2021.09 - 2025.06 (expected), Ph.D. cadidate in Electrical Engineering, advised by Prof. Ying Wu, Northwestern University.
- 2019.09 - 2021.03, M.S. in Electrical Engineering, advised by Prof. Jenq-Neng Hwang, University of Washington.
- 2015.09 - 2019.07, B.S. in Information Engineering, Shanghai Jiao Tong University.
📝 Publications
Active Open-Vocabulary Recognition: Let Intelligent Moving Mitigate CLIP Limitations
Lei Fan, Jianxiong Zhou, Xiaoying Xing, Ying Wu
- Investigate CLIP’s limitations in embodied perception scenarios, emphasizing diverse viewpoints and occlusion degrees.
- Propose an active agent to mitigate CLIP’s limitations, aiming for active open-vocabulary recognition.
Jianxiong Zhou, Ying Wu
- Propose a robust unsupervised anomaly detection method OPFA that can mitigate the influence of contaminated training data and improve the detection of anomalous samples.
- OPFA improves features by contracting normal features and contrasting normal features with outlier features, allowing it to outperform other methods in contaminated data scenarios.
Micro-expression spotting with a novel wavelet convolution magnification network in long videos
Jianxiong Zhou, Ying Wu
- Our approach combines wavelet decomposition and a learnable motion magnification module to enhance the optical flow features of micro-expressions, making them easy to detect.
- We design a selective magnification attention module to suppress background disturbances and highlight MEs. The ablation study and visualization results show that it works as expected.
Jianxiong Zhou, Ying Wu
Video | Supplementary | Poster
- The proposed TFE-DCN has an enlarged receptive field that covers a long temporal span to observe the full dynamics of action instances, making it powerful to capture temporal dependencies between snippets.
- The Modality Enhancement Module can enhance RGB features with enhanced optical flow features to make the overall features suitable for the WTAL task.
-
Hierarchical Pose Classification for Infant Action Analysis and Mental Development Assessment, Jianxiong Zhou, Zhongyu Jiang, Jang-Hee Yoo, and Jenq-Neng Hwang, accepted by International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2021.
-
State‐of‐charge estimation for LiNi0. 6Co0. 2Mn0. 2O2/graphite batteries using the compound method with improved extended Kalman filter and long short‐term memory network, Shuai Xu, Jianxiong Zhou, Fei Zhou, and Yuchen Liu, accepted by International Journal of Energy Research, 2021.
-
State-of-charge estimation of lithium-ion batteries based on improved H infinity filter algorithm and its novel equalization method, Zhenggang Chen, Jianxiong Zhou, Fei Zhou, and Shuai Xu, accepted by Journal of Cleaner Production, 2021.
-
Unfolded coprime planar array for 2D direction of arrival estimation: An aperture-augmented perspective, Wang Zheng, Xiaofeng Zhang, Le Xu, and Jianxiong Zhou, accepted by IEEE Acess, 2018.
💻 Internships
- 2024.06 - 2024.08, Summer Internship, Blinkfire Analytics, Inc., Chicago, US.
- Topic: Be responsible for various engineering tasks including computer vision related tasks. - 2020.06 - 2020.12, Research Intern, Information Processing Lab, Seattle, US.
- Topic: Hierarchical pose classification for infant action analysis and mental development assessment.
- Advisors: Prof. Jenq-Neng Hwang. -
2019.08 - 2019.09, Research Intern, Nanjing Zhenchao Technology Co., Ltd., Nanjing, China.
- Topic: Intelligent audit for documents. -
🎖 Honors and Awards
- 2019.07 Outstanding Graduate of Shanghai Jiao Tong University.
- 2019.04 China Electronic Instrument Scholarship, Shanghai Jiao Tong University
- 2018.04 Honorable Mention, Interdisciplinary Contest in Modeling, COMAP
- 2016.11 China National Scholarship.