• Activity Recognition using Wearables (IMU and EMG sensors) through Deep Learning Networks• Frailty Assessment in the Elderly using Image Representation of IMU Signals and Deep CNN• Gait Events Prediction using CNN, LSTM, GRU, and Hybrid CNN-RNN-based Machine Learning models through a Single Waist-worn Wearable Sensor• Gait-based Human Identification through Machine Learning• Development of 3D Camera Based Foot Measurement and Arch Detection• 3D Reconstruction of Archaeological Pottery Fragments: Pointcloud processing, features extraction and segmentationDetails• Activity Recognition through Deep Learning Networks - Using IMU-based wearable sensors - Using thigh EMG sensors• Deep Learning-based Frailty/CI Multi-Disease Classification using Gait - Feature engineering - Handling with Imbalanced datasets - Multi-label multi-class classification• Gait Event Detection using Single Pelvis Sensor and Attention-based CNN-Bi-RNN Networks - Working with LSTM/GRU/CNN hybrid models• Frailty Assessment in the Elderly using Image Representation of IMU Signals and Deep CNN - Gait Segmentation - Image data generation - Training and testing of CNN models• Gait-based Human Identification through Machine Learning - Temporal feature extraction - Training and testing of machine learning models• Development of 3D Camera Based Foot Shape Assessment - 3D point cloud processing - Foot measurements - Arch detection• 3D Reconstruction of Archaeological Pottery Fragments - 3D point cloud processing - Breakline detection - Feature extraction - 3D point cloud segmentation

Developed several commercial projects related with semiconductor process monitoring and diagnostics under school-industry partnership between Myongji University and Rainbow Corporation. Main tasks included- Real-time windows application development - GUI development- Algorithm development- Real-time process monitoring through sensors- Fault detection and data analysis- Big data visualization and exploration