Metric Embedding Autoencoders for Unsupervised Cross-Dataset Transfer Learning
This addresses a practical challenge in person re-identification systems, but it appears incremental as it builds on existing deep transfer learning approaches.
The paper tackles the problem of cross-dataset transfer learning in person re-identification by proposing a deep convolutional neural network and autoencoder model that separates latent code into metric embedding and nuisance variables, achieving improvements over baseline and competitor models.
Cross-dataset transfer learning is an important problem in person re-identification (Re-ID). Unfortunately, not too many deep transfer Re-ID models exist for realistic settings of practical Re-ID systems. We propose a purely deep transfer Re-ID model consisting of a deep convolutional neural network and an autoencoder. The latent code is divided into metric embedding and nuisance variables. We then utilize an unsupervised training method that does not rely on co-training with non-deep models. Our experiments show improvements over both the baseline and competitors' transfer learning models.