Showing papers by "Anupam Agrawal published in 2010"

A novel human computer interface based on hand gesture recognition using computer vision techniques

Abstract: In daily life, human beings communicate with each other and use broad range of gestures in the process of interaction. Apart of the interpersonal communication, many hours are spent in the interaction with the electronic devices. In the last decade, new classes of devices for accessing information have emerged along with increased connectivity. In parallel to the proliferation of these devices, new interaction styles have been explored. The objective of this paper is to provide a gesture based interface for controlling applications like media player using computer vision techniques. The human computer interface application consists of a central computational module which uses the Principal Component Analysis for gesture images and finds the feature vectors of the gesture and save it into a XML file. The Recognition of the gesture is done by K Nearest Neighbour algorithm. The Training Images are made by cropping the hand gesture from static background by detecting the hand motion using Lucas Kanade Pyramidical Optical Flow algorithm. This hand gesture recognition technique will not only replace the use of mouse to control the media player but also provide different gesture commands which will be useful in controlling the application.

Context-Aware Restricted Geographical Domain Question Answering System

Abstract: With the rapid use of the internet into users’ daily lives, it is imperative to capture location-specific information from local database to satisfy the user’s local information needs [context]. This is especially true in the increasingly local search environments. There are a number of local Web search engines enabling users to find information which is location specific Web content. In this aspect web search engine gives sets of relevant documents related to user query. Thereby user has to scan all searched documents and read line by line to get specific answer. So there is great need to automate the overall process, so that user gets answer in the form of compact text along with map for visualization purpose. This aims at developing geographic domain question answering system [QAS] embedded with mapping abilities where our system will allow users to specify a location addition to the keywords they are searching for as a query. Our system will then return results of the specified location so as to make geographical information retrieval truly context-aware. The testing results show satisfactory performance of the system.

A Vision based Hand Gesture Interface for Controlling VLC Media Player

Abstract: Human Computer Interaction can acquire several advantages with the introduction of different natural forms of device free communication. Gestures are a natural form of actions which we often use in our daily life for interaction, therefore to use it as a communication medium with computers generates a new paradigm of interaction with computers. This paper implements computer vision and gesture recognition techniques and develops a vision based low cost input device for controlling the VLC player through gestures. VLC application consists of a central computational module which uses the Principal Component Analysis for gesture images and finds the feature vectors of the gesture and save it into a XML file. The Recognition of the gesture is done by K Nearest Neighbour algorithm. The theoretical analysis of the approach shows how to do recognition in static background. The Training Images are made by cropping the hand gesture from static background by detecting the hand motion using Lucas Kanade Pyramidical Optical Flow algorithm. This hand gesture recognition technique will not only replace the use of mouse to control the VLC player but also provide different gesture vocabulary which will be useful in controlling the application.

Unusual activity detection for video surveillance

Abstract: ideo surveillance has gained importance in law enforcement, security and military applications The system consists of processing steps such as object detection, movement tracking, and activity monitoring The paper contribution is to present the human activity analysis system that both detect a human with carrying or abandoning an object and segments the object from the human so that it can be tracked Segmentation of objects is done from the background using advance Gaussian mixture model The tracking algorithm considers the human as whole from frame to frame, it does not track the human parts such as limbs Object features such as center of mass, size, and bounding box are used in this paper to estimate a matching between objects in consecutive frames As the object is segmented and tracked, Bayesian inference framework is used for event analysis This system uses a single camera view and unusual activity is detected using the detected objects and object tracking result The operator is notified if an unusual activity is detected

Enabling the interactive display of large medical volume datasets by multiresolution bricking

Abstract: In this paper, we present an approach to interactive out-of-core volume data exploration that has been developed to augment the existing capabilities of the LhpBuilder software, a core component of the European project LHDL ( http://www.biomedtown.org/biomed_town/lhdl ). The requirements relate to importing, accessing, visualizing and extracting a part of a very large volume dataset by interactive visual exploration. Such datasets contain billions of voxels and, therefore, several gigabytes are required just to store them, which quickly surpass the virtual address limit of current 32-bit PC platforms. We have implemented a hierarchical, bricked, partition-based, out-of-core strategy to balance the usage of main and external memories. A new indexing scheme is introduced, which permits the use of a multiresolution bricked volume layout with minimum overhead and also supports fast data compression. Using the hierarchy constructed in a pre-processing step, we generate a coarse approximation that provides a preview using direct volume visualization for large-scale datasets. A user can interactively explore the dataset by specifying a region of interest (ROI), which further generates a much more accurate data representation inside the ROI. If even more precise accuracy is needed inside the ROI, nested ROIs are used. The software has been constructed using the Multimod Application Framework, a VTK-based system; however, the approach can be adopted for the other systems in a straightforward way. Experimental results show that the user can interactively explore large volume datasets such as the Visible Human Male/Female (with file sizes of 3.15/12.03 GB, respectively) on a commodity graphics platform, with ease.

Devanagari character recognition towards natural human-computer interaction

Abstract: Human-computer interaction is a growing research area. There are several ways of interaction with the computer. Handwriting has continued to persist as a means of communication and recording information in the day to day life even with the introduction of new technologies. Due to the growth of technology in India, it becomes important to devise ways that allow people to communicate with computer in Indian languages. Hindi being the national language of India, we present a way to communicate with the computer in Hindi or more precisely, 'Devanagari script'. Due to absence of a global font to represent Devanagari characters, it is important that the computer recognizes the characters written by the user in order to interact with him. The algorithm implemented for character recognition first segments the image containing Devanagari text fed to the software into lines, lines to words and words to characters. The obtained characters are then brought down to a standard size. The Kohonen Neural Network based recognizer then comes into action and recognizes the text character by character and provides the output in Unicode format. The network has been designed with no hidden layer to support quick recognition. Apart from text recognition from an image, we also provided the option to recognize individual handwritten characters drawn using a mouse. Such a system provides keyboard less computer interaction. The technique is implemented using Java. The overall recognition rate for a fixed font machine printed characters is 90.26% and for hand written characters, it is 83.33%.

Hindi - English based cross language Information Retrieval system for Allahabad Museum

Abstract: Cross language information retrieval (CLIR) system is a part of Information Retrieval. It deals with retrieving information in a different language than in which documents are stored. User retrieves the result in the language of his choice. Computing has changed the way to process the data. Our research work develops an application which gives user the flexibility to interact with system in a language of his/her choice. Using this system, a user can enter the query in Hindi or English and retrieve the relevant documents in Hindi or English, though stored documents are in English only. Thus both types of users can get the information through the documents and images. The approach used for this application is to transform the query to English language and then query processing is done. After that retrieval of documents is done, the final output is converted to the desired language. The other way around would be to convert the documents to the query language and then start processing but that would increase the time complexity. In this system, query is in form of keyword in any of the language either Hindi or English. The corpus is of Allahabad museum consisting of images and documents. All the stored documents related to Allahabad museum are in English language. The system successfully handles different types of cross language user queries related to Allahabad Museum.

IMTS- an Interactive Multimodal Tourist-Guide System

Abstract: Today's tourists expect to get easily accessible information about a city which is mostly spatial in nature and the information content is huge. Spatial information or Geographical Information System (GIS) data can be precisely represented and understood by an interactive map. Interaction with GIS data is a complex task and not easily accessible. This paper introduces IMTS (Interactive Multimodal Tourist-Guide System), which is implemented to provide the tourist with such an easy access to get information about a city. IMTS is a scalable map based multimodal tourist guide system. It has navigation control, searching capabilities for hotels, restaurants, attractions, shortest path etc and has more natural multimodal interaction by mouse, keyboard and speech. The outputs are locations, routes on map and contain rich and interactive information like description in form of synthesized speech or text, link and picture. For testing of the system the geospatial data of Allahabad city in India has been used. System integrates Google Map API interfaces, Microsoft .NET Speech SDK and Web 2.0 technologies like JavaScript, AJAX, and Asp.Net with .NET technology. This paper presents technical details of the IMTS system including interaction design and evaluation of the web based tourist guide system.

Proceedings of the First International Conference on Intelligent Interactive Technologies and Multimedia

Abstract: Designing interactive systems is an interdisciplinary field in which computer scientists, engineers, neuro-scientists, radiologists, psychologists, social scientists, design professionals and others play important roles. The goal of this field is to solve real problems in the design and use of technology, making computer controlled and computer-based systems easier to use and more effective for people and organizations. Ease of use fitting the technical level of operator and effectiveness in services are critical to the success of any systems that interact with people, including software systems, home, office and factory appliances, web applications, infotainment systems, medical diagnostics and treatment systems, learning systems etc. Designing interactive systems needs to be human-centered. New paradigms for interacting with computers are being developed which will define the 21st century standards and enable the world to communicate and interact effortlessly and intuitively. Much of today's research in interactive technologies lacks a tight coupling between technology innovation and end-user needs. The development of intelligent interactive systems would need to address the issues related to wide variety of cultures and languages, different levels of literacy and education, new sets of users with different experience, attitudes, expectations and capabilities. The aim of current research in designing intelligent interactive systems is to make the interaction between computers and people smarter. Unlike traditional Human-Machine Interaction (HMI), we are willing more to consider solutions that involve large amounts of knowledge, heuristics, and emerging technologies such as brain-computer interfaces, natural language understanding or gesture recognition etc to serve the purpose effortlessly. The goal of the conference is to bring together researchers working on all aspects of aforesaid and related devices and solutions seamlessly on human-centered design of interactive systems and multimedia technologies. All original contributions will be welcome which address a wide range of theoretical and practical application issues in development of Intelligent Interactive Systems and Multimedia (IITM).