[ICLR2020] Order learning

Kyungsun Lim, Nyeong-Ho Shin, Young-Yoon Lee, and Chang-Su Kim, “Order Learning and Its Application to Age Estimation”

We will present a new concept, called order learning, in ICLR 2020. It is about machine learning of inequality (cf. metric learning is about the learning of equality).

This is our first ICLR paper. Also, this is our first paper to be presented on the African continent. For completeness, we need an Antarctic paper.


[2019] A journal paper on point processing

Congratulations, Seon-Ho, for his first journal paper!
  • Seon-Ho Lee, Han-Ul Kim, and Chang-Su Kim, “ELF-Nets: Deep Learning on Point Clouds Using Extended Laplacian Filter,” to appear in IEEE Access, 2019.

We are developing a liking for this IEEE Access for its quickness and binary decisions. Sadly, traditional review-and-revision doesn’t work anymore. It doesn’t mean that luck-rebuttal-luck of the conference model is perfect. Fortunately, I can retire much earlier than Seon-Ho.

[ICCV2019] Aesthetics assessment and future motion estimation

Two interesting papers will be presented in ICCV, Oct. 2019, Seoul.

* Kyung-Rae Kim, Whan Choi, Yeong Jun Koh, Seong-Gyun Jeong, and Chang-Su Kim, “Instance-Level Future Motion Estimation in a Single Image Based on Ordinal Regression”
* Jun-Tae Lee and Chang-Su Kim, “Image Aesthetic Assessment Based on Pairwise Comparison – A Unified Approach to Score Regression, Binary Classification, and Personalization”

Congratulations Kyung-Rae, Whan, and Jun-Tae! Especially, because Kyung-Rae and Jun-Tae have seen too many summers in Korea University, all MCL members are very happy to hear this good news. A piece of even better news is that Jun-Tae got his degree successfully and will start a new career in Samsung Electronics from this September.

[BMVC2019] A learning algorithm for clustering

Han-Ul Kim, Yeong Jun Koh, and Chang-Su Kim, “Meta Learning for Unsupervised Clustering” in Proc. BMVC, Sept. 2019.

Congratulations Han-Ul for this achievement!

[2019] Yuk ranked 2nd in 2019 DAVIS Challenge on Interactive VOS

Yuk’s interactive video object segmentation algorithm ranked 2nd in the 2019 DAVIS Challenge on Video Object Segmentation. His work has been invited to CVPR workshop 2019.

Yuk Heo, Yeong Jun Koh, Chang-Su Kim, “Interactive Video Object Segmentation Using Sparse-to-Dense Networks,” CVPR Workshop, 2019.

It is a remarkable achievement, considering the complexity of the VOS system and Yuk’s recent change to this topic.  Congratulations, Yuk!  All MCL members wish that this topic would be more interesting than prostate segmentation.