Evaluating bag-of-visual-words representations in scene classification

Abstract

Based on keypoints extracted as salient image patches, an image can be described as a "bag of visual words" and this representation has been used in scene classification. The choice of dimension, selection, and weighting of visual words in this representation is crucial to the classification performance but has not been thoroughly studied in previous work. Given the analogy between this representation and the bag-of-words representation of text documents, we apply techniques used in text categorization, including term weighting, stop word removal, feature selection, to generate image representations that differ in the dimension, selection, and weighting of visual words. The impact of these representation choices to scene classification is studied through extensive experiments on the TRECVID and PASCAL collection. This study provides an empirical basis for designing visual-word representations that are likely to produce superior classification performance.

LoCH: A neighborhood-based multidimensional projection technique for high-dimensional sparse spaces

“…Its instances present 30 dimensions and are classified into two distinct classes, the malignant and benign cancer [39]. The images dataset contains 2117 pictures [40] in three different classes with 32 features extracted using the bag-of-visual features (BoVF) [41] technique. The two_norm and simplex are artificial datasets generated using the mlbench package of R [42].…”

Projection inspector: Assessment and synthesis of multidimensional projections

“…For the sake of illustration, in Fig. 2 we show the layout resulting from the combination of five multidimensional projection techniques, namely, LSP, LAMP, PLMP, SMS, and Hybrid Model (see [26] for a detailed description of those projection methods) when projecting the caltech data set which contains pictures [27] with features extracted using the bag-of-visual features (BoVF) [28] method. Fig.…”

Toward a higher-level visual representation for content-based image retrieval

“…Yang et al in [34] evaluated many frequency weighting schemes which are based on these factors, such as tf-idf weighting, stop word removal, and feature selection. The best weighting scheme in information retrieval does not guarantee good performance in CBIR since the count information can be noisy.…”