A Novel Rate-Scalable Multimedia Service for E-Learning Videos using Content Based Wavelet Compression. Sood, A, Sarthi, D, Pande, A, & Mittal, A In Proceedings of IEEE Annual India Conference 2006, pages 1-6, 2006.
doi  abstract   bibtex   
In this paper, we address the problem of delivery of multimedia services on bandwidth deficient WirelessNetworks for E-learning. We propose a modified continuous rate-scalable wavelet based video compression technique for E-learning lecture videos. The algorithm eradicates spatial and temporal redundancies and exploits the virtues of content based classification for compression. Some of the distinguished features of our algorithm are its excellent subjective performance due to its adaptive behavior to the network conditions and excellent compression due to the classification of the objects in the E-learning lecture video. Also, the compression mechanism maintains high PSNR for prime objects present in the lecture video which further facilitates zooming operations. In this paper, we introduce a noble concept of even and odd frame distinction (EODist). A significant reduction in number of retransmissions of data packets has been observed after integrating EODist with conventional mechanism, which is also a potential advantage of our algorithm in bandwidth deficient wireless network environment. The algorithm exhibits high asymmetry between compressor and de-compressor. The compressor is designed to achieve optimum content based compression using content based wavelet transform and is computationally more complex whereas the decompressor is intentionally designed procedurally simple to support real time de-compression of the compressed data at the client side. The performance of our algorithm is analyzed and its advantages are justified with the help of experimental results
@inproceedings{ ref7,
  title = {A Novel Rate-Scalable Multimedia Service for E-Learning Videos using Content Based Wavelet Compression},
  year = {2006},
  author = {A Sood and D Sarthi and A Pande and A Mittal},
  booktitle = {Proceedings of IEEE Annual India Conference 2006},
  pages = {1-6},
  doi = {10.1109/INDCON.2006.302843},
  abstract = {In this paper, we address the problem of delivery of multimedia services on bandwidth deficient WirelessNetworks for E-learning. We propose a modified continuous rate-scalable wavelet based video compression technique for E-learning lecture videos. The algorithm eradicates spatial and temporal redundancies and exploits the virtues of content based classification for compression. Some of the distinguished features of our algorithm are its excellent subjective performance due to its adaptive behavior to the network conditions and excellent compression due to the classification of the objects in the E-learning lecture video. Also, the compression mechanism maintains high PSNR for prime objects present in the lecture video which further facilitates zooming operations. In this paper, we introduce a noble concept of even and odd frame distinction (EODist). A significant reduction in number of retransmissions of data packets has been observed after integrating EODist with conventional mechanism, which is also a potential advantage of our algorithm in bandwidth deficient wireless network environment. The algorithm exhibits high asymmetry between compressor and de-compressor. The compressor is designed to achieve optimum content based compression using content based wavelet transform and is computationally more complex whereas the decompressor is intentionally designed procedurally simple to support real time de-compression of the compressed data at the client side. The performance of our algorithm is analyzed and its advantages are justified with the help of experimental results},
  keywords = { MultimediaCodingandCommunications }
}

Downloads: 0