20160521-16 欧阳书馨:Forensic Facial Sketch Matching: A time-aware study

报告嘉宾1：欧阳书馨（伦敦玛丽女王大学(SketchX Lab)，北京邮电大学）

报告时间：2016年5月25日（星期三）晚20:00（北京时间）

报告题目：Forensic Facial Sketch Matching: A time-aware study

主持人：  马占宇（北京邮电大学）

报告摘要：We investigate whether it is possible to improve the performance of automated facial forensic sketch matching by learning from examples of facial forgetting over time. Forensic facial sketch recognition is a key capability for law enforcement, but remains an unsolved problem. It is extremely challenging because there are three distinct contributors to the domain gap between forensic sketches and photos: The well-studied sketch-photo modality gap, and the less studied gaps due to (i) the forgetting process of the eye-witness and (ii) their inability to elucidate their memory. Unlike previous studied solving all three factors by one model, we tried to tackle this problem by modeling those factors with different models. We introduced a database of 400 forensic sketches created at different time-delays. Based on this database, we utilize multi-task learning models to model the sketches across time-delays and modalities. We also conducted the experiments with large-mugshot database (10,030). The result outperforms previous method with same scale of data.

参考文献：

[1] ForgetMeNot: Memory-Aware Forensic Facial Sketch Matching, CVPR, 2016

[2] A Survey on Heterogeneous face recognition: sketch, near-infrared, 2D-3D and low-high resolution, IVC, under revision, 2016.