Word Spotting as a Service: An Unsupervised and Segmentation-Free Framework for Handwritten Documents

Word spotting strategies employed in historical handwritten documents face many challenges due to variation in the writing style and intense degradation. In this paper, a new method that permits efficient and effective word spotting in handwritten documents is presented that relies upon document-oriented local features that take into account information around representative keypoints and a matching process that incorporates a spatial context in a local proximity search without using any training data. The method relies on a document-oriented keypoint and feature extraction, along with a fast feature matching method. This enables the corresponding methodological pipeline to be both effectively and efficiently employed in the cloud so that word spotting can be realised as a service in modern mobile devices. The effectiveness and efficiency of the proposed method in terms of its matching accuracy, along with its fast retrieval time, respectively, are shown after a consistent evaluation of several historical handwritten datasets.

Weighted proximity search

Proximity search is an iterative method to solve complex mathematical programming problems. At each iteration, the objective function of the problem at hand is replaced by the Hamming distance function to a given solution, and a cutoff constraint is added to impose that any new obtained solution improves the objective function value. A mixed integer programming solver is used to find a feasible solution to this modified problem, yielding an improved solution to the original problem. This paper introduces the concept of weighted Hamming distance that allows to design a new method called weighted proximity search. In this new distance function, low weights are associated with the variables whose value in the current solution is promising to change in order to find an improved solution, while high weights are assigned to variables that are expected to remain unchanged. The weights help to distinguish between alternative solutions in the neighborhood of the current solution, and provide guidance to the solver when trying to locate an improved solution. Several strategies to determine weights are presented, including both static and dynamic strategies. The proposed weighted proximity search is compared with the classic proximity search on instances from three optimization problems: the p-median problem, the set covering problem, and the stochastic lot-sizing problem. The obtained results show that a suitable choice of weights allows the weighted proximity search to obtain better solutions, for 75$$\%$$ % of the cases, than the ones obtained by using proximity search and for 96$$\%$$ % of the cases the solutions are better than the ones obtained by running a commercial solver with a time limit.

Engineering Graph Features via Network Functional Blocks

Graph is a prevalent data structure that enables many predictive tasks. How to engineer graph features is a fundamental question. Our concept is to go beyond nodes and edges, and explore richer structures (e.g., paths, subgraphs) for graph feature engineering. We call such richer structures as network functional blocks, because each structure serves as a network building block but with some different functionality. We use semantic proximity search as an example application to share our recent work on exploiting different granularities of network functional blocks. We show that network functional blocks are effective, and they can be useful for a wide range of applications.