Ding-Jie Chen

PostDoc, Academia Sinica

Email: djchen.tw [at] gmail [dot] com

[DBLP] [Google Scholar] [CV] [GitHub]

I am currently a postdoctoral fellow at IIS Computer Vision Lab, advised by Dr. Tyng-Luh Liu. I received my Ph.D. from the Department of Computer Science, National Tsing Hua University, advised by Professor Hwann-Tzong Chen and Professor Long-Wen Chang. Prior to joining National Tsing Hua University, I got my Master's Degree from Tatung University, advised by Professor Shang-Lin Hsieh. 

I am passionate about solving emerging computer vision problems. My current research lies in aggregating computer vision and deep learning to design impactful techniques. Specifically, I am of interest in tackling challenging computer vision problems, including various segmentation and detection tasks involving scenarios such as multi-modal learning, few-shot learning, attention mechanism, and anomaly detection. If you are interested in my research or have any use cases that you want to share, please feel free to contact me!

Publications - Journal

[3]   SwipeCut: Interactive Segmentation via Seed Grouping

[2]   Toward a Unified Scheme for Fast Interactive Segmentation

[1]   Interactive 1-bit Feedback Segmentation Using Transductive Inference

Publications - Conference, Workshop

[24]  Contrastive Feature Decoupling for Weakly-Supervised Disease Detection.

[23]  One-Shot Action Detection via Attention Zooming In

[22]  Aggregating Bilateral Attention for Few-Shot Instance Localization

[21]  Self-Supervised Sparse Representation for Video Anomaly Detection

[20]   Contextual Proposal Network for Action Localization

[19]   Learning Unsupervised Metaformer for Anomaly Detection

[18]   Adaptive Image Transformer for One-Shot Object Detection

[17]   Referring Image Segmentation via Language-Driven Attention

[16]   Temporal Action Proposal Generation via Deep Feature Enhancement

[15]   See-Through-Text Grouping for Referring Image Segmentation

[14]   Instance-Level Meta Normalization

[13]   Unsupervised Meta-learning of Figure-Ground Segmentation via Imitating Visual Effects

[12]  A2A: Attention to Attention Reasoning for Movie Question Answering

[11]   Tap and Shoot Segmentation

[10]   Video Object Segmentation via Cellular Automata Refinement

[09]   Detecting Nonexistent Pedestrians

[08]   Video Segmentation via Boundary-Aware Flow

[07]   Interactive Segmentation from 1-Bit Feedback

[06]   Fast Defocus Map Estimation

[05]   Unsupervised Image Co-segmentation based on Cooperative Game

[04]   Modified Soft-decision Adaptive Interpolation by an Evolutionary Game

[03]   Matching based Image Co-segmentation

[02]   Video Object Cosegmentation

[01]   An RST-resilient Image Copyright Protection Scheme based on the Invariant Domain and Image Secret Sharing