Skip to content

Academic

🖋️ Handwritten Devanagari Script Recognition (2015)

  • Objective: Build machine learning models to recognize handwritten characters in Devanagari script.
  • Features: Provided pre-processed datasets, training scripts, and evaluation tools.
  • Outcome: Enabled researchers and developers to explore character recognition efficiently.
  • Academic Context: Bachelor's Project, GEC Goa.

📡 Total Electron Content (TEC) as a Seismic Precursor (2017)

  • Objective: Use TEC data and Multi-Fractal Detrended Fluctuation Analysis (MFDFA) to identify seismic activities.
  • Key Insights: Analyzed fractal patterns in ionospheric variations for early earthquake detection.
  • Academic Context: Master's Project, GEC Goa & IIT Bombay.

🔄 Hybrid Evolutionary Optimization (HEO) of GANs (2019)

  • Objective: Enhance GAN training using evolutionary algorithms.
  • Key Contributions:
  • HEO: Combined genetic algorithms with backpropagation for improved performance.
  • Hybrid GAN: Leveraged MFDFA in the discriminator to generate images faster and more efficiently.
  • 2D MFDFA: Integrated multifractal analysis for temporal and spatial data, refining GAN learning.
  • Outcome: Improved convergence and high-quality image generation.
  • Academic Context: PhD Project, NIT Goa.