Completed:
Biomedical Signal Processing:
- Development of a GUI for Neurosurgeons using time frequency analysis
of EEG signals.
- Time frequency & non-linear dynamic analysis of Heart Rate
Variability for detection of impending heart attacks.
- Development of an algorithm to detect differences between vector
probability distributions to find epileptic patterns in EEG. (
Report
)
- Time-series modeling using adaptive competitive mixture of experts
(MOE) & threshold auto regressive (TAR) models towards segmentation
of EEG data.
Speech Processing:
- Implementation of 8kbps speech codec using CS-ACELP Algorithm
according to G.729 standard specified by ITU-T.
- A chaotic approach to Speech Recognition. (
Proposal
Report
)
VLSI:
- Test of a VGA capture FPGA.
- Implementation of b3zs encoder and decoder.
- Design and simulation of a 48 bit SRAM using HP 0.5um process
in SPECTRE.
Adaptive Signal
Processing:
- Speech enhancement using adaptive noise cancellation & inverse
modeling with RLS, LMS & other gradient descent algorithms.
Neural Networks:
- Design of a Neural Network classifier to separate sustained notes
played in guitar.
- Design of a Time Delayed Neural Network to predict chaotic time-series.
Current:
Biomedical Signal
Processing:
- Design of an automated algorithm to detect epileptic seizures from
EEG data.