Sanjeev Satheesh

Text Detection and Character Recognition in Scene Images with Unsupervised Feature Learning

Reading text from photographs is a challenging problem that has received a significant amount of attention. Two key components of most systems are (i) text detection from images and (ii) character recognition, and many recent methods have been proposed to design better feature representations and models for both. In this paper, we apply methods recently developed in machine learning -- specifically, large-scale algorithms for learning the features automatically from unlabeled data -- and show that they allow us to construct highly effective classifiers for both detection and recognition to be used in a high accuracy end-to-end system.

Temporal classification of events in cricket videos

Video search today uses the metadata surrounding the video, ignoring its semantic content. Over the years, a lot of research has gone into indexing and browsing of sports video content. In this work, we present a novel approach for classification of events in cricket videos and thus, summarize its visual content. The proposed method segments a cricket video into shots and identifies the visual content in them. Using sequential pattern mining and support vector machine, we classify the sequence of shots into four events, namely RUN, FOUR, SIX and OUT. The cricket video is then summarized based on user-supplied parameters. The performance of the system has been tested on a number of cricket video clips and was found to have an accuracy of 87.8%.