Sai Krishna Charan Dara
Home
Education
Experience
Publications
ResearchProjects
Course Projects
Research Blog
AdvancedCPP
View My LinkedIn Profile
Lets Connect
Course Projects
- Computer Science Projects
- Core ECE Projects
- ALU Design using Full Swing GDI technique
- Time Frequency Analysis of Non Stationary Electrocardiogram Signals Using Hilbert Huang Transform
- OFDM and OFDMA
- Adaptive Modulation Applications
- Wavelet based denoising of ECG Signal
- 6T-SRAM Memory Array
- Insertion Sort on FPGA
- Square Wave Generator
- Class-D Power Amplifier
- Neyman Pearson Detector for Variance Estimation
Computer Science Projects
Interactive C Shell
- Course: - Operating Systems
- Developed a user-interactive shell in C, implementing the major features of GNU/Linux shell like piping, redirecting, handling background and foreground processes
File Sharing Protocol between client and server
- Course: - Communication Networks
- File sharing protocol is created between client and server using socket programming which had functionalities like Indexed searching, File Hashing (using MD5 checksum), File transfer (UDP and TCP) and caching.
Style Transfer for Headshot Portraits
- Course: - Digital Image Processing
- Headshot portraits are a popular subject in photography but to achieve a compelling visual style requires advanced skills that a casual photographer will not have.
- Presented a technique to transfer the style of an example head-shot photo onto a new one. This can allow one to easily reproduce the look of renowned artists. At the core of our approach is a new multiscale technique to robustly transfer the local statistics of an example portrait onto a new one. This technique matches properties such as the local contrast and the overall lighting direction while being tolerant to the unavoidable differences between the faces of two different people.
-
Our proposed method output looks like:-
Universal Style Transfer via Feature Transforms
- Course: - Computer Vision
- Given a pair of examples the content and style image, universal style transfer aims to synthesize an image that preserves some notion of the content but carries characteristics of the style. The key challenge is how to extract effective representations of the style and then match it in the content image.
- Objective is to develop a universal style transfer approach with a decent visual quality and efficiency.
- Used VGG-19 as the feature extractor. Our algorithm pipeline is in 3 steps. 1) Reconstruction (Encoder and Decoder) to calculate weights. (Used Coco DataSet). 2) Whitening Coloring Transform 3) Multi Level Stylization.
- Patch Transfer is also done as an additional study.
Face Classification
- Course: - Statistical Methods in AI
- Performed data reduction techniques like K-PCA, LDA and build classifers using Multi Layer Perceptron(MLP) and Support Vector Machines(SVM) on popular datasets for distinguishing image classes.
Multi Dimensional Divide and Conquer
- Course: - Operating Systems
- Used this paradigm to solve domination and closest point problems in k-dimensional space and performed complexity analysis.
- Algorithms analyzed are All-points ECDF problem, ECDF Searching Problem, Maxima Finding, Maxima Searching, Range Searching, Fixed-Radius Near Neighbors, Nearest Neighbours.
Core ECE Projects
ALU-Design-using-Full-Swing-GDI-technique
- Course: - Digital VLSI Design
- Desinged a 4-bit Arithmetic Logical Unit in LTSpice using Full Swing GDI technique with optimized Area, Speed and Transistor count.
- Also talked about various logic design techniques and we justify why GDI (Gate Diffusion Input) technique is the best compared to others.
OFDM and OFDMA
- Course: - Communication Networks
- End to end modelling of OFDM and OFDMA is studied and compared with other existed multiplexing and multiple access schemes
Adaptive Modulation Applications
- Course: - Wireless Communications
- Studied different adaptive modulation techniques for various scenarios and analyzed how BER and average spectral efficiency improve with adaption.
- Also analyzed Varialbe Rate - Variable Power MQAM, Adaptive techniques in Fast and Slow Fading, Use of Adaptive techniques for Coding Issues and Capacity and discussed present case scenario and technology (5G) where adaptive technquies are used.
- Various simulations are done using MATLAB.
Time Frequency Analysis of Non Stationary Electrocardiogram Signals Using Hilbert Huang Transform
- Course: - Time Frequency Analysis
- TFAnalysis of ECG Signals using HHTransform and comparing our results with other transforms like STFT,Wavelet Transform and Pseudo Wigner Ville Distribution
- Intrisic Mode Functions (IMF’s) are generated using the algorithm EMD (Empirical mode decomposition).
- Tweaked some parameters in HHT and analysed their effects
- Various simulations are done using MATLAB.
Wavelet based denoising of ECG Signal
- Course: - Digital Signal Processing
- Detrended and then denoised the ECG signal using wavelet denoising technique making QRS complex more distinct and identified peaks and valleys of denoised ECG signal.
- Our main aim of project is to remove those noises and detect QRS complex of ECG signal.
- This project is a modified version of Pan-Tompkins algorithm which detects QRS complexes in ECG signals.
-
Our proposed method output looks like:-
6T-SRAM Memory Array
- Course: - Digital VLSI Design
- Modelled Parasitic capacitances and analysis of Noise Margin, Power, Delay, Rise Time and Fall Time is done.
- Analysis is done using Cadence-Virtuoso
- Design Details (Links of pdfs)
Insertion Sort on FPGA
- Course: - Embedded Hardware Design
- Implemented accelerated Insertion sort on FPGA (Zedboard Zync-7000).
- Included Parallelism and blocks like BRAM , FIFO. Xilinx-Vivado is the software used.
- Also analyzed power consumption and complexity of algorithm compared to normal processor.
Square Wave Generator
- Course: - Linear Electronic Circuits
- Implemented Transistor level Design of Square wave generator on Cadence with minimum power consumption and MOSFET’s being in subthreshold region of operation.
- Building blocks of design are Current Source, 2-Stage Amplifier (acts as comparator), D-FlipFlop and Buffer.
- Above blocks are designed using Cadence-Virtuoso in Transistor level.
Class-D Power Amplifier
- Course: - Electronic Workshop
- Implemented Class D power amplifier with efficiency around 80-90% on breadboard.
Neyman Pearson Detector for Variance Estimation
- Course: - Multi Variate Analysis
- Used Neyman Pearson Detector for detection of Aircrafts using the variance of the signal received.