Fine-grained Similarity Research Articles

A medical image retrieval method based on texture block coding tree

Content-based medical image retrieval (CBMIR) has been widely studied for computer aided diagnosis. Accurate and comprehensive retrieval results are effective to facilitate diagnosis and treatment. Texture is one of the most important features used in CBMIR. Most of existing methods utilize the distances between matching point pairs for texture similarity measurement. However, the distance based similarity measurements are of low tolerance to slight texture shifts, which result in an excessive sensitivity. Furthermore, with the increase of the number of texture points, their time complexity is in explosive growth. In this paper, a new medical image retrieval model is presented based on an iterative texture block coding tree. The corresponding methods for coarse-grained and fine-grained similarity matching are also proposed. Moreover, a multi-level index structure is designed to enhance the retrieval efficiency. Experimental results show that, our methods are of high efficiency and appropriate tolerance on slight shifts, and achieve a relative better retrieval performance in comparison of other existing methods.

Positioning of online gambling and gaming products from a consumer perspective: A blurring of perceived boundaries

Traditionally disparate markets and products, gambling activities and games have begun to overlap in many ways, particularly in the online marketplace. Consumers appear to be moving between these activities, indicating perception of similarities in products. The hybrid nature of some activities has resulted in gambling regulators considering whether they need to act with regards to certain games. Typically, regulatory classifications and taxonomies have been centred on objective features of product offerings, based on the perspective of producers and technology. However, consumer perspective is relevant to defining markets and understanding the potential migratory relationship between products. This research aims to gather an indepth understanding about the multifacetal market space of online gambling and gaming from a consumer perspective. A large-scale survey (n = 1000 respondents) assesses perceived similarities between 16 gaming and 9 gambling products by means of holistic triadic comparisons. Multidimensional scaling provides evidence for enduring market boundaries between online gambling and gaming products, but indicates several fine-grained similarities on the level of specific products. Products encountering elements of skill, planning, consideration, and achievements over time were perceived differently from those that are more playful and less realistic with immediate outcomes. Consumers did see connections between games and some gambling products, which may explain the joint usage of and migration between products. Insights are thus relevant for regulators, consumers, and professionals in both gambling and gaming industries as market boundaries have become blurred.

Automatic action annotation in weakly labeled videos

Manual spatio-temporal annotation of human actions in videos is laborious, requires several annotators and contains human biases. In this paper, we present a weakly supervised approach to automatically obtain spatio-temporal annotations of an actor in action videos. We first obtain a large number of action proposals in each video. To capture a few most representative action proposals in each video and evade processing thousands of them, we rank them using optical flow and saliency in a 3D-MRF based framework and select a few proposals using MAP based proposal subset selection method. We demonstrate that this ranking preserves the high quality action proposals. Several such proposals are generated for each video of the same action. Our next challenge is to iteratively select one proposal from each video so that all proposals are globally consistent. We formulate this as Generalized Maximum Clique Graph problem using shape, global and fine-grained similarity of proposals across the videos. The output of our method is the most action representative proposals from each video. Our method can also annotate multiple instances of the same action in a video. We have validated our approach on three challenging action datasets: UCF-Sport, sub-JHMDB and THUMOS13 and have obtained promising results compared to several baseline methods. Moreover, action detection experiments using annotations obtained by our method and several baselines demonstrate the superiority of our approach.

Exploiting the complementary strengths of multi-layer CNN features for image retrieval

Deep convolutional neural networks have demonstrated breakthrough accuracies for image classification. A series of feature extractors learned from CNN have been used in other computer vision tasks. However, CNN features of different layers aim to encode different-level information. High-layer features care more about semantic information but less detail information, while low-layer features contain more detail information but suffer from the problem of background clutter and semantic ambiguity. We propose to exploit complementary strengths of different layers in a simple but effective way. A mapping function is designed to highlight the effectiveness of low-layer similarity, when measuring fine-grained similarity between query image and its nearest neighbors with similar semantic. Extensive experiments show that our method can achieve competitive performance on popular retrieval benchmarks. Extensive experiments show that the proposed method outperforms the features extracted from single layers and their direct concatenations. Meanwhile, our method achieves competitive performance on popular retrieval benchmarks.

Exploratory analysis of textual data streams

In this paper, we address exploratory analysis of textual data streams and we propose a bootstrapping process based on a combination of keyword similarity and clustering techniques to: (i) classify documents into fine-grained similarity clusters, based on keyword commonalities; (ii) aggregate similar clusters into larger document collections sharing a richer, more user-prominent keyword set that we call topic; (iii) assimilate newly extracted topics of current bootstrapping cycle with existing topics resulting from previous bootstrapping cycles, by linking similar topics of different time periods, if any, to highlight topic trends and evolution. An analysis framework is also defined enabling the topic-based exploration of the underlying textual data stream according to a thematic perspective and a temporal perspective. The bootstrapping process is evaluated on a real data stream of about 330.000 newspaper articles about politics published by the New York Times from Jan 1st 1900 to Dec 31st 2015.

Automatically Mining Facets for Queries from Their Search Results

We address the problem of finding query facets which are multiple groups of words or phrases that explain and summarize the content covered by a query. We assume that the important aspects of a query are usually presented and repeated in the query’s top retrieved documents in the style of lists, and query facets can be mined out by aggregating these significant lists. We propose a systematic solution, which we refer to as QDMiner, to automatically mine query facets by extracting and grouping frequent lists from free text, HTML tags, and repeat regions within top search results. Experimental results show that a large number of lists do exist and useful query facets can be mined by QDMiner. We further analyze the problem of list duplication, and find better query facets can be mined by modeling fine-grained similarities between lists and penalizing the duplicated lists.

Word Similarity Computing Based on Hybrid Hierarchical Structure by HowNet

Word similarity computing is one of the most important and fundamental task in the field of natural language processing. Most of word similarity methods perform well in synonyms, but not well between words whose similarity is vague. It confronts the challenge of how to overcome this problem. An approach is proposed to compute Chinese word similarity based on hybrid hierarchical structure by HowNet to achieve fine-grained similarity results. The experimental results prove that the method has a better effect on computing similarity of synonyms and antonyms including nouns, verbs and adjectives. In addition, it performs well and stably on standard data provided by SemEval 2012.

Social ecology of similarity

Social ecologies shape the way people initiate and maintain social relationships. Settings with much opportunity will lead to more fine-grained similarity among friends; less opportunity leads to less similarity. We compare two ecological contexts—a large, relatively diverse state university versus smaller colleges in the same state—to test the hypothesis that a larger pool of available friendship choices will lead to greater similarity within dyads. Participants in the large campus sample reported substantially more perceived ability to move in and out of relationships compared to participants in the small colleges sample. Dyads were significantly more similar on attitudes, beliefs, and health behaviors in the large campus than in the small colleges sample. Our findings reveal an irony—greater human diversity within an environment leads to less personal diversity within dyads. Local social ecologies create their own “cultures” that affect how human relationships are formed.

Child safety education information

Hashing based cross-modal retrieval has recently made significant progress. But straightforward embedding data from different modalities involving rich semantics into a joint Hamming space will inevitably produce false codes due to the intrinsic modality discrepancy and noises. We present a novel deep Robust Multilevel Semantic Hashing (RMSH) for more accurate multi-label cross-modal retrieval. It seeks to preserve fine-grained similarity among data with rich semantics,i.e., multi-label, while explicitly require distances between dissimilar points to be larger than a specific value for strong robustness. For this, we give an effective bound of this value based on the information coding-theoretic analysis, and the above goals are embodied into a margin-adaptive triplet loss. Furthermore, we introduce pseudo-codes via fusing multiple hash codes to explore seldom-seen semantics, alleviating the sparsity problem of similarity information. Experiments on three benchmarks show the validity of the derived bounds, and our method achieves state-of-the-art performance.

Words and voices: episodic traces in spoken word identification and recognition memory.

Most theories of spoken word identification assume that variable speech signals are matched to canonical representations in memory. To achieve this, idiosyncratic voice details are first normalized, allowing direct comparison of the input to the lexicon. This investigation assessed both explicit and implicit memory for spoken words as a function of speakers' voices, delays between study and test, and levels of processing. In 2 experiments, voice attributes of spoken words were clearly retained in memory. Moreover, listeners were sensitive to fine-grained similarity between 1st and 2nd presentations of different-voice words, but only when words were initially encoded at relatively shallow levels of processing. The results suggest that episodic memory traces of spoken words retain the surface details typically considered as noise in perceptual systems.