3 years ago, we found out that the performance of existing CPUs was not enough to deliver real-time processing of scenes for autonomous vehicles. ISLab started working with Nvidia CUDA technology which led to a successful speedup of our algorithms. At that time, there were not so many researchers working with GPUs, but we think that the development of the detection algorithm is not enough. The algorithm should be applicable to the real systems so the society benefits from the new inventions.
The link to the news: https://news.developer.nvidia.com/vision-smoke-detection-using-low-cost-surveillance-cameras/
ISLAB contributed 2 papers to the ICGIP2017 :
1. Qing Tang et al., Traffic Sign Classification with Dataset Augmentation and Convolutional Neural Network
2.Yang Yu et al., Lane Detection based on Color Probability Model and Fuzzy Clustering
ICT-ROBOT 2017 was held at the Kagoshima University, Korimoto Campus, Japan from Nov. 25 to Nov. 26, 2017. This conference was organized with the SICE Kyushu branch annual conference.
ISLab contributes 3 papers:
- Dong-Wook Seo, Hyun-Deok Kang, and Kang-Hyun Jo, “Multi-layer superpixel cut for stereo matching”
- Beom-Seok Kim, Gi-Seok Sohn, and Kang-Hyun Jo, “Illegal Trash Littering Person Detection for Intelligent Surveillance System”
- Youlkyeong Lee and Kang-Hyun Jo, “Local Disparity Computation with SIFT and Belief Propagation”
The 43rd annual conference of the IEEE Industrial electronic society (IECON) was held in the china national convention center, Beijing, China from 29 October until 1 November 2017. The conference venue is located nearby the Olympic village which is located around 30 km from the Beijing international airport. In this year, it attracts more than 2400 of submitted papers. Among them, around 1400 papers are presented in either oral presentation or poster presentation.
ISLAB contribute 2 papers:
1. Laksono Kurnianggoro and Kang-Hyun Jo, Identification of Pedestrian Attributes using Deep Network
2. Ajmal Shahbaz and Kang-Hyun Jo, Optimal Background Modeling for Cluttered Scenes
The 2017 Thirteenth International Conference on Intelligent Computing (ICIC 2017) was held during August 7-10, 2017, in Liverpool, UK. The conference will be financially supported by the Natural Science Foundation of China (Natural Science Foundation of China). The theme for this conference is Advanced Intelligent Computing Methodologies and Applications. Original contributions related to this theme are especially solicited, including theories, methodologies, and applications in science and technology. Topics covering industrial issues/applications and academic research into intelligent computing will be welcome.
ISLab contributed with 1 paper:
Wahyono and Kang-Hyun Jo, Human Carrying Baggage Classification Using Transfer Learning on Convolution Neural Network with Direction Attribute, ICIC 2017 D.-S. Huang et al. (Eds.): ICIC 2017, Part I, LNCS 10361, Liverpool,UK, Aug 7, 2017 pp.717-724.
ISLab successfully organized 10th International Conference on Human System Interaction (HSI 2017) at University of Ulsan (UOU). It was held under the banner of IEEE Industrial Electronics Society (IES) Ulsan Chapter, co-sponsored by UOU and BK21+. Delegates from 15 different countries joined the conference. ISLab contributed with 7 papers:
1) Sowmya Kasturi et al., Classification of Human Fall in Top Viewed Kinect Depth Images using Binary Support Vector Machine
2) Hoang Van Thanh et al., An Improved Method for 3D Shape Estimation Using Active Shape Model
3) Dongwook Seo et al., Multi-Layer Superpixel-Based MeshStereo for Accurate Stereo Matching
4) Ajmal Shahbaz et al., Sterile Zone Monitoring with Human Verification
5) Alexander Filonenko et al., Comparative Study of Modern Convolutional Neural Networks for Smoke Detection on Image Data
6) Laksono Kurnianggoro et al., Object Classification for LIDAR Data using Encoded Features
7) Sanghee Lee et al., Strategy for Automatic Person Indexing and Retrieval System in News Interview Video Sequences
Heartiest congratulations to Professor Kang-Hyun Jo and team!
The 15th International Conference on Industrial Informatics (INDIN`2017) is an international conference for researchers and academics share ideas and experiences surrounding frontier technologies, breakthroughs, innovative solutions, research results, as well as innovation initiatives related to industrial informatics and their applications. In this year, this conference was held in Emden, Germany, between the 24th and the 26th July 2017.
ISLAB contribute 2 papers:
1. Sanghee Lee, and Kanghyun Jo, Automatic Person Information Extraction Using Overlay Text in Television News Interview Videos, INDIN 2017, Emden, Germany, July 25, 2017.
2. Van-Thanh Hoang, Van-Dung Hoang, and Kang-Hyun Jo, An Improved Method for 3D Shape Estimation Using Cascade of Neural Networks, INDIN 2017, Emden, Germany, July 26, 2017.
Our undergraduate school student Participated in KSIC 2017.(The Korean Society of Industry Convergence)
A total of 6 teams participated.
1) Su-Jin Seol et. al, Audio and Video Stimulated Mood Light
2) Jae-Hee Lee et. al, IoT based Secured Window Control
3) Young Joo Kim et. al, Efficient Energy and Eco-Friendly Humidity Controller
4) Dong-Hoon Lee et. al, Robot cleaner
5) Dea-Youb Kim et. al, Design and Implementation of Missing Child Prevention System
Using GPS Module and Bluetooth Communication Device
6) Yun-Jeong Choi et. al, Emotional Mood light based on the Weather
ISIE 2017 was held in Edinburgh, United Kingdom from 19-21 June. ISLab had one contribution:
Ajmal et al., Optimal Color Space based Probabilistic Foreground Detector for Video Surveillance System
ISLab Participated in ICROS 2017 with 3 contributions.
Kyu-Jin Park"s paper was awarded best paper award by Institue of Control, Robotics and Systems.
1) Kyu-Jin Park et. al, Method of Improving Clearness through Contrast of Background and Object using Histogram Deformation
2) Beom-Seok Kim et. al, Detection of Thrower and Garbage Using the Analysis of Pedestrian Trajectory and Baggage4
3) Gi-Seok Sohn et. al, Determination of the Status and Kind of Fruits according to Color Distribution
ACIIDS is an international conference for researches in the filed of intelligent information and database systems. In this year, this conference was held during 3-5 April 2017 in Kaga, Japan.
ISLAB contribute 2 papers:
1. Ajmal Shahbaz, Laksono Kurnianggoro, Wahyono, Kang-Hyun Jo, Recent Advances in the Field of Foreground Detection: An Overview, ACIIDS 2017, Kanazawa, Japan, Apr 3, 2017.
2. Laksono Kurnianggoro, Wahyono, and Kang-Hyun Jo, Improving Traffic Sign Recognition Using Low Dimensional Features, ACIIDS 2017, Kanazawa, Japan, Apr 3, 2017.