Publications

Journal papers: 

• 1.  Sharma A, Katlaa R, Kaur G, Jayagopi DB. Full-page handwriting recognition and automated essay scoring for in-the-wild essays. Multimedia Tools and Applications. 2023 Mar 13:1-24.

  2. Sharma, A. and Jayagopi, D.B., 2021. Towards efficient unconstrained handwriting recognition using Dilated Temporal Convolution Network. Expert Systems with Applications, 164, p.114004.

  3. Annapurna Sharma, Hoon Jae lee, Wan- Young Chung, “ Principal Component Analysis based Ambulatory monitoring of Elderly”, The Journal of Korea institute of maritime information & communication sciences, Vol. 12. No. 11, pp.2105-2110, Nov. 2008.

International Conference Papers: 

• 1. Sharma, Annapurna, Rahul Ambati, and Dinesh Babu Jayagopi. "Towards Faster Offline Handwriting Recognition Using Temporal Convolution Networks." In Computer Vision, Pattern Recognition, Image Processing, and Graphics: 7th National Conference, NCVPRIPG 2019, Hubballi, India, December 22–24, 2019, Revised Selected Papers 7, pp. 344-354. Springer Singapore, 2020.

  2. A. Sharma, D. Jayagopi. Handwritten Essay Grading on Mobiles using MDLSTM Model and Word Embeddings, published in the 11th Indian Conference on Computer Vision, Graphics and Image Processing (ICVGIP), Hyderabad, India, Dec 2018.

  3. A. Sharma and D. Jayagopi. Automated Grading of Handwritten Essays, published in  16th International Conference on Frontiers in Handwriting Recognition (ICFHR-2018), to be held August 5-8, 2018 in Niagara Falls, USA.

  4. Annapurna Sharma,  Amit Purwar, Young-Dong Lee, Young-Sook Lee, Wan-Young Chung, "Frequency based classification of activities using accelerometer data",  Multisensor Fusion and Integration for Intelligent Systems, 2008. MFI 2008. IEEE International Conference on , vol., no., pp.150-153, 20-22 Aug. 2008, Seoul, Korea.

  5. Wan-Young Chung, Amit Purwar,  Annapurna Sharma, "Frequency domain approach for activity classification using accelerometer", Engineering in Medicine and Biology Society, 2008. EMBS 2008. 30th Annual International Conference of the IEEE , vol., no., pp.1120-1123, 20-25 Aug. 2008, Vancouver, Canada .

  6. Annapurna Sharma, Young-Dong Lee, Wan-Young Chung, “High Accuracy Human Activity Monitoring using Neural Network”, The 2008 international conference on Convergence and Hybrid Information Technology. ICCIT 2008, vol. no. 1, pp.430-435, 11-13 Nov. 2008, Pusan, Korea. 

  7.  Annapurna Sharma, Hakimjon Zaynidinov, Hoon Jae Lee, “Development and Modelling of High-Efficiency Computing Structure for Digital Signal Processing”, International Conference on Multimedia, Signal Processing and Communication Technologies, IMPACT-2009. Aligarh, India. 

Domestic Conference Papers: 

• 1.  Annapurna Sharma, Amit Purwar, Wan young Chung, “Automatic activity classification from accelerometer data in Sensor network environment”, Sensor society, Korea Institute of Information Security and Cryptography(KIISC 2008), pp.71-75,  Feb 2008, Busan, Korea.  

  2. Annapurna Sharma, Young-Dong Lee, Wan-Young Chung, “Human activity classification using neural network”, The Korean Institute of Maritime Information & Communication Science(KIMICS), vol.12,  no.1, pp.229-232, May 2008, Gwangju, Korea.

  3. Annapurna Sharma, Hoon Jae Lee, Wan-Young Chung, “Neural Network design for Ambulatory monitoring of elderly”, The Korean Institute of Maritime Information & Communication Science(KIMICS), vol.12,  no.2, pp.265-269,  Oct 2008, Pusan, Korea.

Master’s Thesis: “A PCA-based Neural Network Classifier Approach for Wireless Physical Activity Monitoring”

Thesis Supervisors:

Prof. Hoon Jae Lee 

Prof Hyotaek Lim

Prof Sug Hyon Tae

Prof. Wan Young Chung

Thesis Abstract:  The encroachment of technology in daily life style has fostered a growing interest for the ubiquitous healthcare systems. Activity monitoring from remote location using wireless technology is one of the application gaining ample of attention due to wide variety of concerns. Several systems have been proposed by different research groups. The accuracy of classification is one of the key issues if such a system is proposed for a healthcare application. This dissertation proposes an offline algorithm to be used to classify the human physical activities using the data received from a single tri-axial accelerometer sensor node in the WSN environment. Complete system architecture for the proposed method is shown. The system uses transmission of the acceleration anomalies associated with the user physical activity to a remote wireless sensor node via an IEEE 802.15.4 compliant radio module. The detailed analysis and selection of various parameters using the 3-axis acceleration data is shown with the separation of body and gravity acceleration components. The work shows an approach in designing a high accuracy PCA-based NN classifier for classifying very simple activities of daily living i.e. rest, walk and run with a consideration of several T- and F- domain features. The signal processing, feature set preparation and the reduction using PCA are proposed along with neural network. The results show an overall classification rate of 94.88% which is sufficiently high as compared to other methods and existing researches.