Goal: Measure the deformable change of aorta along the time. 

Achievement: Propose a novel method to decompose the Jacobian into radial and longitudinal directions; Apply the DL-based segmentation on CT; Automate the centerline extraction. 

Keywords: Registration, Directional Jacobian, CNN for Segmentation

>>> See overview here

Goal: Perform weak-supervised segmentation for skin lesion images (vitiligo). 

Achievement: Propose a dataset; Propose a weak-supervised segmentation for vitiligo images, and achieves IoU at 71.4 which is comparable with SOTA ( e.g. PRM, SEC-CRF.)

Keywords: CNN, Saliency Propagation,  Weakly-supervised Segmentation

>>> See more results: [slides] [talk] [paper]

Goal: Remove specular highlight in facial images

Achievement: Our method generates highlight-free images with high-quality details and fewer artifacts by adopting CGAN and multi-scale discriminators.

Keywords:  Highlight Removal,  Facial Image,  CGAN

>>> See abstract and overview here

Goal: Estimate 6D pose from single RGB image.

Achievement:  Our method achieves BOP score at 61.4 on YCB-V using only RGB images, which is 10.6 higher than original Pix2Pose’s performance.

Keywords:  6D pose,  PnP,  U-net,  single RGB

>>> See more results: [report]

Goal: Apply key frame extraction and video condensation to surveillance videos.

Achievement: Implement a video condensation system from scratch using C++; Designed a novel motion-volume based keyframe extraction method; Second Prize of National-level SRTP(Student Research Training Program)

Keyword: K-means, Motion-volume, Graph-cut, Tracking ...

>>> See more results here...