Seunghyeok Back

I am a Ph.D. candidate, advised by Professor Kyoobin Lee, at the School of Integrated Technology (Robotics Program), Gwangju Institute of Science and Technology (GIST), South Korea. I received a B.S. degree in mechanical engineering from GIST, South Korea in 2018.

My research interests are in robot vision for unseen object manipulation. I focused on the robotic perception and manipulation of unseen objects via deep neural networks and Sim2Real transfer. I have developed and integrated various robotic systems, such as vision-based grasping, furniture assembly from IKEA instruction, and human-robot collaboration in an unstructured environment. My research topic also includes the health care application of deep learning in biomedical signals and the image domain.

news

Feb 1, 2022	Our work on unseen object amodal instance segmentation have been accepted at ICRA 2022!
Jan 27, 2022	🏅 Won Bronze prize at 28th Humantech Paper Award by Samsung Electronics! ($5,000)
Dec 21, 2021	🏅 Won Outstanding Paper Award at ICROS-KROC Honam Conference.
May 24, 2021	🏅 Won Outstanding Paper Award at KROC 2021.

selected publications

ICRA

Unseen Object Amodal Instance Segmentation via Hierarchical Occlusion Modeling

Back, Seunghyeok, Lee, Joosoon, Kim, Taewon, Noh, Sangjun, Kang, Raeyoung, Bak, Seongho, and Lee, Kyoobin

Accepted at 2022 IEEE International Conference on Robotics and Automation (ICRA 2022) 2022

Abs PDF Supp Code Website

Instance-aware segmentation of unseen objects is essential for a robotic system in an unstructured environment. Although previous works achieved encouraging results, they were limited to segmenting the only visible regions of unseen objects. For robotic manipulation in a cluttered scene, amodal perception is required to handle the occluded objects behind others. This paper addresses Unseen Object Amodal Instance Segmentation (UOAIS) to detect 1) visible masks, 2) amodal masks, and 3) occlusions on unseen object instances. For this, we propose a Hierarchical Occlusion Modeling (HOM) scheme designed to reason about the occlusion by assigning a hierarchy to a feature fusion and prediction order. We evaluated our method on three benchmarks (tabletop, indoors, and bin environments) and achieved state-of-the-art (SOTA) performance. Robot demos for picking up occluded objects, codes, and datasets are available at https://sites.google.com/view/uoais.
BSPC

Intra-and inter-epoch temporal context network (IITNet) using sub-epoch features for automatic sleep scoring on raw single-channel EEG

Seo, Hogeon, Back, Seunghyeok, Lee, Seongju, Park, Deokhwan, Kim, Tae, and Lee, Kyoobin (first co-author)

Biomedical Signal Processing and Control 2020

Abs PDF Supp

A deep learning model, named IITNet, is proposed to learn intra- and inter-epoch temporal contexts from raw single-channel EEG for automatic sleep scoring. To classify the sleep stage from half-minute EEG, called an epoch, sleep experts investigate sleep-related events and consider the transition rules between the found events. Similarly, IITNet extracts representative features at a sub-epoch level by a residual neural network and captures intra- and inter-epoch temporal contexts from the sequence of the features via bidirectional LSTM. The performance was investigated for three datasets as the sequence length () increased from one to ten. IITNet achieved the comparable performance with other state-of-the-art results. The best accuracy, MF1, and Cohen’s kappa () were 83.9%, 77.6%, 0.78 for SleepEDF ( = 10), 86.5%, 80.7%, 0.80 for MASS ( = 9), and 86.7%, 79.8%, 0.81 for SHHS ( = 10), respectively. Even though using four epochs, the performance was still comparable. Compared to using a single epoch, on average, accuracy and MF1 increased by 2.48%p and 4.90%p and F1 of N1, N2, and REM increased by 16.1%p, 1.50%p, and 6.42%p, respectively. Above four epochs, the performance improvement was not significant. The results support that considering the latest two-minute raw single-channel EEG can be a reasonable choice for sleep scoring via deep neural networks with efficiency and reliability. Furthermore, the experiments with the baselines showed that introducing intra-epoch temporal context learning with a deep residual network contributes to the improvement in the overall performance and has the positive synergy effect with the inter-epoch temporal context learning.
ICIP

Segmenting unseen industrial components in a heavy clutter using rgb-d fusion and synthetic data

Back, Seunghyeok, Kim, Jongwon, Kang, Raeyoung, Choi, Seungjun, and Lee, Kyoobin

In 2020 IEEE International Conference on Image Processing (ICIP) 2020

Abs PDF Supp Code

Segmentation of unseen industrial parts is essential for autonomous industrial systems. However, industrial components are texture-less, reflective, and often found in cluttered and unstructured environments with heavy occlusion, which makes it more challenging to deal with unseen objects. To tackle this problem, we present a synthetic data generation pipeline that randomizes textures via domain randomization to focus on the shape information. In addition, we propose an RGB-D Fusion Mask R-CNN with a confidence map estimator, which exploits reliable depth information in multiple feature levels. We transferred the trained model to real-world scenarios and evaluated its performance by making comparisons with baselines and ablation studies. We demonstrate that our methods, which use only synthetic data, could be effective solutions for unseen industrial components segmentation.