Recurrent Transformer Networks For Semantic Correspondence

Authors:
Seungryong Kim Yonsei University
Stephen Lin Microsoft Research
SANG RYUL JEON Yonsei University
Dongbo Min Ewha Womans University
Kwanghoon Sohn Yonsei Univ.

Introduction:

The authors present recurrent transformer networks (RTNs) for obtaining dense correspondences between semantically similar images.

Abstract:

We present recurrent transformer networks (RTNs) for obtaining dense correspondences between semantically similar images. Our networks accomplish this through an iterative process of estimating spatial transformations between the input images and using these transformations to generate aligned convolutional activations. By directly estimating the transformations between an image pair, rather than employing spatial transformer networks to independently normalize each individual image, we show that greater accuracy can be achieved. This process is conducted in a recursive manner to refine both the transformation estimates and the feature representations. In addition, a technique is presented for weakly-supervised training of RTNs that is based on a proposed classification loss. With RTNs, state-of-the-art performance is attained on several benchmarks for semantic correspondence.

You may want to know: