Hey everybody,
Today I'll share a work that we published back in 2017 and was originally prepared as a CVPR 2016 when I was a PhD student at ETH Zürich.
Although dated, the approach could easily be modernized by the use of new architectures. The idea is very simple and was novel at the time of publication; although I've seen it used in many many papers afterwards. Use the detector of bounding boxes to guide the segmentation; and then utilize a hashing technique on the given patch to retrieve nearest neighbors. You can then retrieve, for instance, top-n and merge, where n could change depending on your problem at-hand.
The abstract goes like this; again, please *do* cite it if you use it as a source of inspiration.
We propose a novel approach to address the problem of Simultaneous Detection and Segmentation introduced in [Hariharan et al. 2014]. Using the hierarchical structures first presented in [Arbeláez et al. 2011] we use an efficient and accurate procedure that exploits the feature information of the hierarchy using Locality Sensitive Hashing. We build on recent work that utilizes convolutional neural networks to detect bounding boxes in an image [Ren et al. 2015] and then use the top similar hierarchical region that best fits each bounding box after hashing, we call this approach C&Z Segmentation. We then refine our final segmentation results by automatic hierarchical pruning. C&Z Segmentation introduces a train-free alternative to Hypercolumns [Hariharan et al. 2015]. We conduct extensive experiments on PASCAL VOC 2012 segmentation dataset, showing that C&Z gives competitive state-of-the-art segmentations of objects.
Segmentation of Objects by Hashing.
De Curtò, De Zarzà, Smola and Gool.
Today I'll share a work that we published back in 2017 and was originally prepared as a CVPR 2016 when I was a PhD student at ETH Zürich.
Although dated, the approach could easily be modernized by the use of new architectures. The idea is very simple and was novel at the time of publication; although I've seen it used in many many papers afterwards. Use the detector of bounding boxes to guide the segmentation; and then utilize a hashing technique on the given patch to retrieve nearest neighbors. You can then retrieve, for instance, top-n and merge, where n could change depending on your problem at-hand.
The abstract goes like this; again, please *do* cite it if you use it as a source of inspiration.
We propose a novel approach to address the problem of Simultaneous Detection and Segmentation introduced in [Hariharan et al. 2014]. Using the hierarchical structures first presented in [Arbeláez et al. 2011] we use an efficient and accurate procedure that exploits the feature information of the hierarchy using Locality Sensitive Hashing. We build on recent work that utilizes convolutional neural networks to detect bounding boxes in an image [Ren et al. 2015] and then use the top similar hierarchical region that best fits each bounding box after hashing, we call this approach C&Z Segmentation. We then refine our final segmentation results by automatic hierarchical pruning. C&Z Segmentation introduces a train-free alternative to Hypercolumns [Hariharan et al. 2015]. We conduct extensive experiments on PASCAL VOC 2012 segmentation dataset, showing that C&Z gives competitive state-of-the-art segmentations of objects.
Segmentation of Objects by Hashing.
De Curtò, De Zarzà, Smola and Gool.
De Curtò i DíAz.