Adaptive Cascade Deep Convolutional Neural Networks for face alignment

摘要: Abstract Deep convolutional network cascade has been successfully applied for face alignment. The configuration of each network, including the selecting strategy local patches training and input range patches, is crucial achieving desired performance. In this paper, we propose an adaptive framework, termed Adaptive Cascade Convolutional Neural Networks (ACDCNN) which adjusts structure adaptively. Gaussian distribution utilized to bridge successive networks. Extensive experiments demonstrate that our proposed ACDCNN achieves state-of-the-art in accuracy, but with reduced model complexity increased robustness.