Chained Predictions Using Convolutional Neural Networks
arXiv:1605.02346
Abstract
In this paper, we present an adaptation of the sequence-to-sequence model for structured output prediction in vision tasks. In this model the output variables for a given input are predicted sequentially using neural networks. The prediction for each output variable depends not only on the input but also on the previously predicted output variables. The model is applied to spatial localization tasks and uses convolutional neural networks (CNNs) for processing input images and a multi-scale deconvolutional architecture for making spatial predictions at each time step. We explore the impact of weight sharing with a recurrent connection matrix between consecutive predictions, and compare it to a formulation where these weights are not tied. Untied weights are particularly suited for problems with a fixed sized structure, where different classes of output are predicted in different steps. We show that chained predictions achieve top performing results on human pose estimation from single images and videos.
arXiv commentin submission to EECV 2016
References in corpus (8)
- Sequence to Sequence Learning with Neural Networks
- Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
- DeepPose: Human Pose Estimation via Deep Neural Networks
- Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
- Rethinking the Inception Architecture for Computer Vision
- Stacked Hourglass Networks for Human Pose Estimation
- Listen, Attend and Spell
- Order Matters: Sequence to sequence for sets