Egocentric Vision for Visual Market Basket Analysis
This paper introduces a new application scenario for egocentric vision: Visual Market Basket Analysis (VMBA). The main goal in the proposed application domain is the understanding of customers behaviours in retails from videos acquired with cameras mounte
- PDF / 4,009,761 Bytes
- 14 Pages / 439.37 x 666.142 pts Page_size
- 47 Downloads / 195 Views
Department of Mathematics and Computer Science, University of Catania, Catania, Italy [email protected] 2 Centro Studi S.r.l., Buccino, SA, Italy
Abstract. This paper introduces a new application scenario for egocentric vision: Visual Market Basket Analysis (VMBA). The main goal in the proposed application domain is the understanding of customers behaviours in retails from videos acquired with cameras mounted on shopping carts (which we call narrative carts). To properly study the problem and to set the first VMBA challenge, we introduce the VMBA15 dataset. The dataset is composed by 15 different egocentric videos acquired with narrative carts during users shopping in a retail. The frames of each video have been labelled by considering 8 possible behaviours of the carts. The considered cart’s behaviours reflect the behaviour of the customers from the beginning (cart picking) to the end (cart releasing) of their shopping in a retail. The inferred information related to the time of stops of the carts within the retail, or to the shops at cash desks could be coupled with classic Market Basket Analysis information (i.e., receipts) to help retailers in a better management of spaces and marketing strategies. To benchmark the proposed problem on the introduced dataset we have considered classic visual and audio descriptors in order to represent video frames at each instant. Classification has been performed exploiting the Directed Acyclic Graph SVM learning architecture. Experiments pointed out that an accuracy of more than 93 % can be obtained on the 8 considered classes.
1
Introduction and Motivations
Egocentric vision is a new emerging area in Computer Vision [1,2]. By exploiting wearable devices it is possible to collect hours of videos that can be processed to obtain a log of the monitored scenarios. Different papers on egocentric vision applications have been published in the recent literature. The main tasks addressed in this area are related to scene recognition [3], motion understanding [4], objects and actions recognition [5–8], 3D reconstruction [9,10] and summarization [11,12]. Among the others, context aware computing is an important research area for egocentric (first-person) vision domain [3,13,14]. Temporal segmentation of Egocentric Vision is also fundamental to understand the behavior of the users wearing a camera [4,15]. Recently, the retail scenario has c Springer International Publishing Switzerland 2016 G. Hua and H. J´ egou (Eds.): ECCV 2016 Workshops, Part I, LNCS 9913, pp. 518–531, 2016. DOI: 10.1007/978-3-319-46604-0 37
Egocentric Vision for Visual Market Basket Analysis
519
Fig. 1. Information useful for VMBA.
become of particular interest for applications related to the geo-localization of the user’s positions and the reconstruction of the spaces [16]. In the retail context, one of the possible developments of interest concerns the monitoring of the paths of customers, thereby enabling to carry out an analysis of their behaviors. Nowadays customers monitoring is partially employed by
Data Loading...
