Convolutional aggregation of local evidence for large pose face alignment

British Machine Vision Conference 2016 Adrian BulatGeorgios Tzimiropoulos

Methods for unconstrained face alignment must satisfy two requirements: they must not rely on accurate initialisation/face detection and they should perform equally well for the whole spectrum of facial poses. To the best of our knowledge, there are no methods meeting these requirements to satisfactory extent, and in this paper, we propose Convolutional Aggregation of Local Evidence (CALE), a Convolutional Neural Network (CNN) architecture particularly designed for addressing both of them... (read more)

PDF

Code


No code implementations yet. Submit your code now
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK LEADERBOARD
Face Alignment AFLW-PIFA (21 points) Face alignment NME 2.63% # 1
Face Alignment AFLW-PIFA (34 points) Face alignment NME 2.96% # 1