Project information

• Created a Gaussian-based SLAM framework that integrates object detection for enhanced localization and mapping in dynamic environments.
• Applied Gaussian state estimation for real-time object tracking, enabling precise position and movement predictions over sequential frames.
• Integrated YOLOv7 for object detection, extracting bounding box information from Gaussian-splatted images, and rendered 3D bounding boxes on point clouds using depth data for enhanced spatial visualization.
• Project Report: Link