Study on the correlation between color and taste of beauty tea infusion and the pivotal contributing compounds based on UV-visible spectroscopy, taste equivalent quantification and metabolite analysis.

This study utilized a colorimeter to determine the color values of 23 beauty tea (BT) samples, the color and the taste characteristics were also quantitatively described through ultraviolet-visible (UV-Vis) spectroscopy and taste equivalent quantification. Furthermore, metabolomic analysis was conducted by using ultra-high-performance liquid chromatography-mass spectrometry (UPLC-MS). Correlation analysis was employed to preliminarily identify the compounds that contribute to the color and taste of BT infusion. Finally, the contributing compounds were further determined through verification experiment. The results showed that within a certain range, as the color of BT infusion deepened, the taste became stronger, more bitter and astringent, while on the contrary, it became sweeter and mellower. Theaflavins, kaempferol, astragalin, and 5-p-coumaroylquinic acid influenced both the color and taste of the BT infusion. Gallic acid was also determined as a contributor to the color. This study provides new insights into research on tea quality in infusion color and taste aspects.

Robust ImageGraph: Rank-Level Feature Fusion for Image Search.

Recently, feature fusion has demonstrated its effectiveness in image search. However, bad features and inappropriate parameters usually bring about false positive images, i.e., outliers, leading to inferior performance. Therefore, a major challenge of fusion scheme is how to be robust to outliers. Towards this goal, this paper proposes a rank-level framework for robust feature fusion. First, we define Rank Distance to measure the relevance of images at rank level. Based on it, Bayes similarity is introduced to evaluate the retrieval quality of individual features, through which true matches tend to obtain higher weight than outliers. Then, we construct the directed ImageGraph to encode the relationship of images. Each image is connected to its K nearest neighbors with an edge, and the edge is weighted by Bayes similarity. Multiple rank lists resulted from different methods are merged via ImageGraph. Furthermore, on the fused ImageGraph, local ranking is performed to re-order the initial rank lists. It aims at local optimization, and thus is more robust to global outliers. Extensive experiments on four benchmark data sets validate the effectiveness of our method. Besides, the proposed method outperforms two popular fusion schemes, and the results are competitive to the state-of-the-art.