Image resizing by reconstruction from deep features

Dov Danon, Moab Arar, Daniel Cohen-Or, Ariel Shamir

Research output: Contribution to journalArticlepeer-review


Traditional image resizing methods usually work in pixel space and use various saliency measures. The challenge is to adjust the image shape while trying to preserve important content. In this paper we perform image resizing in feature space using the deep layers of a neural network containing rich important semantic information. We directly adjust the image feature maps, extracted from a pre-trained classification network, and reconstruct the resized image using neural-network based optimization. This novel approach leverages the hierarchical encoding of the network, and in particular, the high-level discriminative power of its deeper layers, that can recognize semantic regions and objects, thereby allowing maintenance of their aspect ratios. Our use of reconstruction from deep features results in less noticeable artifacts than use of imagespace resizing operators. We evaluate our method on benchmarks, compare it to alternative approaches, and demonstrate its strengths on challenging images.

Original languageEnglish
Pages (from-to)453-466
Number of pages14
JournalComputational Visual Media
Issue number4
StatePublished - Dec 2021


  • deep seam carving
  • image resizing
  • image retargeting
  • reconstruction


Dive into the research topics of 'Image resizing by reconstruction from deep features'. Together they form a unique fingerprint.

Cite this