Prof. P. P. Vaidyanathan, California Institute of Technology (Caltech), USA|
Title of the Talk: Srinivasa Ramanujan and Digital Signal Processing
The great Indian mathematician Srinivasa Ramanujan introduced a summation in 1918, called the Ramanujan-sum. For many years this summation was used by mathematicians to prove important results in number theory. In recent years, some researchers have found applications of this sum in digital signal processing, especially in identifying periodic components of signals buried in noise. In our recent work we have generalized the Ramanujan- sum decomposition in several directions, and this has opened up some new theory as well as applications in the representation and identification of structured signals such as periodic signals. Many beautiful properties are enjoyed by such representations, thanks to the genius and vision of Ramanujan. Our new developments include Ramanujan dictionaries, Ramanujan filter banks, and the Ramanujan MUSIC algorithm, and applications of these in the study of DNA and protein sequences among others. In this talk we will give an overview of the theory and show some applications.
|Prof. Lina J. Karam, Arizona State University (ASU), USA|
Title of the Talk: Generative Sensing: Transforming Unreliable Sensor Data for Reliable Recognition
This talk introduces a deep learning enabled generative sensing framework which integrates low-end sensors with computational intelligence to attain a high recognition accuracy on par with that attained with high-end sensors. The proposed generative sensing framework aims at transforming low-end, low-quality sensor data into higher quality sensor data in terms of achieved classification accuracy. The low-end data can be transformed into higher quality data of the same modality or into data of another modality. Different from existing methods for image generation, the proposed framework is based on discriminative models and targets to maximize the recognition accuracy rather than a similarity measure. This is achieved through the introduction of selective feature regeneration in a deep neural network (DNN).