Gait Energy Image Reconstruction from Degraded Gait Cycle Using Deep Learning

Maryam Babaee, Linwei Li, Gerhard Rigoll; Proceedings of the European Conference on Computer Vision (ECCV) Workshops, 2018, pp. 0-0

Gait energy image (GEI) is considered as an effective gait representation for gait-based human identification. In gait recognition, normally, GEI is computed from one full gait cycle. However in many circumstances, such a full gait cycle might not be available due to occlusion. Thus, the GEI is not complete, giving a rise to degrading gait identification rate. In this paper, we address this issue by proposing a novel method to reconstruct a complete GEI from a few frames of gait cycle. To do so, we propose a deep learning-based approach to transform incomplete GEI to the corresponding complete GEI obtained from a full gait cycle. More precisely, this transformation is done gradually by training several fully convolutional networks independently and then combining these as a uniform model. Experimental results on a large public gait dataset, namely OULP demonstrate the validity of the proposed method for gait identification when dealing with very incomplete gait cycles.