A comparison of dense region detectors for image search and fine grained classification
Download as DOCX, PDF0 likes163 views
The document evaluates various methods for dense region detection in image classification and search, proposing alternatives like superpixels, edges, and Zernike filters. Results indicate that these methods outperform traditional dense detectors in most scenarios, enhancing state-of-the-art performance in image retrieval and fine-grained classification. Additionally, it is highlighted that extracting patches along edges improves accuracy for blob and superpixel methods.
A comparison of dense region detectors for image search and fine grained classification
- 1. A COMPARISON OF DENSE REGION DETECTORS FORIMAGE SEARCH AND FINE-GRAINED CLASSIFICATION ABSTRACT We consider a pipeline for image classification or search based on coding approaches like bag of words or Fisher vectors. In this context, the most common approach is to extract the image patches regularly in a dense manner on several scales. This paper proposes and evaluates alternative choices to extract patches densely. Beyond simple strategies derived from regular interest region detectors, we propose approaches based on superpixels, edges, and a bank of Zernike filters used as detectors. The different approaches are evaluated on recent image retrieval and fine-grained classification benchmarks. Our results show that the regular dense detector is outperformed by other methods in most situations, leading us to improve the state-of-the-art in comparable setups on standard retrieval and fined-grained benchmarks. As a byproduct of our study, we show that existing methods for blob and superpixel extraction achieve high accuracy if the patches are extracted along the edges and not around the detected regions.