MDPV: metric distance permutation vocabulary

Abstract

Sub-image content-based similarity search forms an important operation in current image archives since it provides users with images that contain a query image as their part. Such a search can conveniently be implemented using the bag-of-features model. Its integral part is a construction of visual vocabulary. Most existing algorithms to create a visual vocabulary suffer from high computational (e.g. k-means) or supervisor-guidance (e.g. visual-bit classifier, or sparse coding) requirements. In this paper, we propose a novel approach to visual vocabulary construction called metric distance permutation vocabulary. It is based on permutations of metric distances to create compact visual words. Its major advantage over prior techniques is time and space efficiency of vocabulary construction and quantization process during querying, while achieving comparable or even better effectiveness (query result quality). Moreover, this basic concept is extended to combine more independent permutations. Both the proposals are experimented on well-known real-world data-sets and compared to other state-of-the-art techniques.

Motion Words: A Text-Like Representation of 3D Skeleton Sequences

“…The image quantization strategies have evolved from basic k-means clustering used in [21], through cluster hierarchies [14], approximate k-means [16], to recent deep neural-network approaches [24]. The influence of the border problem can be reduced using a weighted combination of the nearest visual words for each feature [16], or by a consensus voting of multiple independent vocabularies [6].…”

Distribution Entropy Boosted VLAD for Image Retrieval

“…Although the inverted index mechanism [8] and hierarchical clustering methods, such as vocabulary tree [5], can improve memory usage and index efficiency, the encoding consumption cannot be improved effectively. Recently, MDPV (Metric Distance Permutation Vocabulary) uses permutations of metric distances to create compact visual words to attain time and space efficiency of vocabulary construction [11]. Moreover, BoW adopts a simple counting method for each cluster to build the final representation.…”