Multiple object cues for high performance vector quantization

作者：

Highlights：

• A multi-cue representation is proposed for high performance vector quantization.

• Keypoint detection using differential entropy is introduced for efficient sampling.

• Shape representation is improved using the proposed co-occurrence statistics.

• We report high accuracy on Caltech-101 dataset using the multi-cue representation.

• We report best results on Flickr-101 dataset using the multi-cue representation.

摘要

•A multi-cue representation is proposed for high performance vector quantization.•Keypoint detection using differential entropy is introduced for efficient sampling.•Shape representation is improved using the proposed co-occurrence statistics.•We report high accuracy on Caltech-101 dataset using the multi-cue representation.•We report best results on Flickr-101 dataset using the multi-cue representation.

论文关键词：Log-polar transform,Object classification,Visual cues,Bag-of-words model,Flickr-101 dataset,Caltech-101 dataset

论文评审过程：Received 25 April 2016, Revised 8 December 2016, Accepted 21 February 2017, Available online 22 February 2017, Version of Record 6 March 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.02.024