Research Interests

The theoritical section of my B.Sc thesis is uploaded here:

thesis_complete_29.09.09.pdf (4,5 MB)

The topic of my M. Sc thesis was acuostic echo and noise cancellation using adaptive filters and optimization algorithms. The thesis title was, "Acoustic Echo and Noise Cancellation Schemes Using Time and Frequency Domain Adaptive Techniques". I have published some results in several conference papers and submitted to some journals. Dr. Shaikh Anowarul Fattah , Associate Professor, Dept. of EEE, BUET supervised my thesis works. The work differs from other conventional acoustic echo cancellation (AEC) methods based on adaptive filter algorithms in the following criteria:

1. It is formulated for single channel. Single Channel Adaptive Filtering is difficult because there is no seperate channel for reference signal.
2. The objective function is diffrent from traditional gradient based algorithms.
3.  An evolutionary algorithm, namely, the particle swarm optimization algorithm (PSO) is employed to solve the problem of AEC. No good reference on PSO based AEC can be found.
4. The problem of AEC in noise is handled to simultaneously cancel out echo and noise in the single channel scenario.
5. Multi-step stopping criteria is developed based on speech properties to control the update process of the adaptive filter to improve performance.

Extensive experimentation is carried out on real life speech signals corrupted by echoes using the proposed single channel LMS algorithm with and without the multi-step update constraints. It is found that the performance of former one, the controlled LMS algorithm, is far better than that of the later one in terms of (a) the average echo return loss enhancement (ERLE) in dB and (b) the difference between input- and output-signal to distortion ratio (SDR) in dB.

I learned the OpenCV, the image processing toolbox of MATLAB and I joined in a web based discussion group led by Dr. Md. Atiqur Rahman Ahad, author of "Computer Vision and Action Recognition" available from Springer. I am currently working on action recognition using optical flow, Random Sample Consensus (RANSAC), statistical signal processing, spectral analysis techniques, motion history image (MHI), motion energy image (MEI), histogram of oriented gradient (HOG) etc. for feature extraction and  PCA based classification.

As a teacher, I supervised a number of projects done by my students such as

1. Fuzzy Logic Driven Adaptive PID Controller for PWM Based Buck Converter
2.  Series Active Power Filter Implementation Using P-Q Theory 
3. Shunt Active Power Filter Implementation Using P-Q Theory
4. voice-controlled wheelchair
5. FPGA implementation of real-time gesture recognition algorithm
6. webcam to dot-matrix video converter
7. Digital Oscilloscope with special DSP functions
8. Line following Robots
9. Tutorial on Microcontrollers (episodes 1, 2, 3, 4) (for the students of Bangladesh)

and many more.

A detailed list of my publications can be found in my CV.