object tracking

"结果表明，在所研究的静态物体映射场景中，SODA-CitrON 在 F1 分数、位置 RMSE、MOTP 和 MOTA 方面持续优于比较方法。"

"The research focuses on unsupervised multi-CCTV dynamic object tracking."

"The research focuses on hand-object tracking."

"The research focuses on event-based multi-object tracking."

"The system achieves 96pJ/Frame/Pixel and 61pJ/Event performance."

"Deep Learning is utilized for tracking custom objects in video."