Pose-Normalized Image Generation for Person Re-identification
arXiv:1712.02225
Abstract
Person Re-identification (re-id) faces two major challenges: the lack of cross-view paired training data and learning discriminative identity-sensitive and view-invariant features in the presence of large pose variations. In this work, we address both problems by proposing a novel deep person image generation model for synthesizing realistic person images conditional on the pose. The model is based on a generative adversarial network (GAN) designed specifically for pose normalization in re-id, thus termed pose-normalization GAN (PN-GAN). With the synthesized images, we can learn a new type of deep re-id feature free of the influence of pose variations. We show that this feature is strong on its own and complementary to features learned with the original images. Importantly, under the transfer learning setting, we show that our model generalizes well to any new re-id dataset without the need for collecting any training data for model fine-tuning. The model thus has the potential to make re-id model truly scalable.
arXiv comment10 pages, 5 figures
References in corpus (19)
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
- Improving Person Re-identification by Attribute and Identity Learning
- A Discriminatively Learned CNN Embedding for Person Re-identification
- Pose Guided Person Image Generation
- Autoencoding beyond pixels using a learned similarity metric
- Deep Representation Learning with Part Loss for Person Re-Identification
- GLAD: Global-Local-Alignment Descriptor for Pedestrian Retrieval
- Pose Invariant Embedding for Deep Person Re-identification
- SVDNet for Pedestrian Retrieval
- Attribute2Image: Conditional Image Generation from Visual Attributes
- HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis
- Scalable Person Re-identification on Supervised Smoothed Manifold
- Learning to Recognize Pedestrian Attribute
- Attribute Recognition by Joint Recurrent Learning of Context and Correlation
- Semi-Latent GAN: Learning to generate and modify facial images from attributes
- Person Re-identification in the Wild
- Nonlinear Local Metric Learning for Person Re-identification
- Weakly-supervised Learning of Mid-level Features for Pedestrian Attribute Recognition and Localization