DeepSketchHashing

alt tag

This is the repository for reproducing some key results for our paper

to be presented on CVPR 2017 spotlight section. This work focuses on fast sketch-based image retrieval (SBIR) using binary codes.

Prerequisites

To produce the binary codes described in the paper, one needs to install Caffe beforehand.

The mid-level Sketch-Token representation is required for training the model. The codes can be found here. Please refer to the following papers for more details.

Models

We provide several pretrained models on two datasets with their respective deploy files. You may try to use any of these models to produce hash code for image-sketch matching.

Sketchy Dataset (Extended)

  • 64 bits
  • 128 bits
  • (UPDATED 4 AUG 2017!)The extra image data mentioned in the paper can be found here (NEW). The previous uploaded image data are wrong (apologize for this).
  • 64 bits
  • 128 bits
  • (UPDATED 21 AUG 2017!)The images of TU for our experiments can be found here.