The field of computer vision is moving towards more accurate and robust tracking and representation of 3D objects and shapes. Recent developments have focused on combining different modalities, such as camera and LiDAR data, to improve tracking efficiency and accuracy. Additionally, there is a growing interest in using physics-based models and contrastive learning to enhance the representation and tracking of 3D shapes. Notable papers in this area include:

• Contrail-to-Flight Attribution Using Ground Visible Cameras and Flight Surveillance Data, which introduces a modular framework for attributing contrails to their source flight.
• Transformed Multi-view 3D Shape Features with Contrastive Learning, which demonstrates the effectiveness of combining Vision Transformers with contrastive objectives for 3D representation learning.
• FutrTrack: A Camera-LiDAR Fusion Transformer for 3D Multiple Object Tracking, which proposes a modular camera-LiDAR multi-object tracking framework that builds on existing 3D detectors.
• Physics-Guided Fusion for Robust 3D Tracking of Fast Moving Small Objects, which combines deep learning-based detection with physics-based tracking to overcome the limitations of existing approaches.
• Radar-Camera Fused Multi-Object Tracking: Online Calibration and Common Feature, which presents a Multi-Object Tracking framework that fuses radar and camera data to enhance tracking efficiency.