Research

My research is primarily focused on harnessing the power of deep learning and machine learning algorithms to address real-world challenges. By leveraging advanced techniques such as graph neural networks (GNNs) and time-series modeling, my work provides innovative solutions in the analysis of pedestrian and vehicle trajectory predictions, as well as human motion analysis. These areas demand extensive spatial-temporal analysis to understand and predict dynamic behaviors.

Spatial Modeling From the spatial modeling perspective, my research is dedicated to developing graph-based architectures that represent diverse data sets, including pedestrians, vehicles, and human skeletal structures. Utilizing cutting-edge GNN algorithms, I strive to elucidate the intricate relationships within this data, facilitating more accurate predictions and richer analytical insights. The application of graph-based approaches allows for a nuanced understanding of spatial interactions that traditional methods might overlook.

Temporal Modeling Regarding temporal modeling, my work involves employing recurrent neural networks (RNNs), Temporal Convolutional Networks (TCNs), and Transformers to model temporal dependencies accurately. This aspect of my research is crucial for predicting future states and behaviors by analyzing how patterns evolve over time, providing a robust framework for temporal analysis in motion-related datasets.