多外观模型的鲁棒人脸跟踪

为了解决在真实场景中进行视觉人脸跟踪时不同挑战之间的外观匹配问题，提出了一种多外观模型的人脸跟踪算法。该算法利用多个具有长期和短期外观记忆的外观模型进行有效的人脸跟踪，对变形、旋转、尺度和光照变化表现出鲁棒性。同时利用先检测后跟踪算法的优点，通过使用人脸检测器来处理人脸的剧烈外观变化，检测器也有助于在漂移过程中重新初始化所提算法。最后提出了一种加权分数级融合策略，通过在可能的人脸位置生成的候选人脸来获得融合值最高的人脸跟踪输出。实验结果证明，该跟踪器在自启动时表现出色，性能优于许多先进的跟踪器。

Robust face tracking with multiple appearance model

To solve the problem of appearance matching across different challenges in visual face tracking in real-world scenarios,a robust face tracking algorithm with multiple appearance model is proposed.The algorithm utilizes multiple appearance model with its long-term and short-term appearance memory for efficient face tracking.It demonstrates robustness to deformation,rotation,scale,and illumination variation.Meanwhile,it capitalizes on the advantages of the tracking-by-detection,by using a face detector that tackles drastic scale appearance change of a face.The detector also helps to reinitialize the proposed algorithm during drift.A weighted score-level fusion strategy is proposed to obtain the face tracking output with the highest fusion score by generating candidates around possible face locations.The tracker shows impressive performance when initiated automatically and outperformes many state-of-the-art trackers.