Hi! I’m Aftaab Ahamed

Personal Projects :

• Classification of spectra of -OH and -RCOOH groups using CNNs :

  Automatically classify spectra of the functional groups by training a CNN

  • Tools & technologies used :TensorFlow, numpy, openCV, pandas, selenium, pyautogui, matplotlib

  • Utilised selenium and pyautogui to collect training data by scraping the SDBS spectra database

  • Preprocess spectra graph data using openCV

  • Trained TensorFlow CNN classification model that yielded 85% accuracy on validation dataset

• Voice Assistant/Chatbot :

  Developed voice assistant using google-gemini api

  • Tools & technologies used :Python, Gemini LLM api, gTTs, picovoice api, NLTK

  • Built a Voice assistant using gemini LLM api for generating responses

  • Wake word detection implemented using picovoice api and text to speech using gTTs library

  • Calculated cosine similarity of sentences using NLTK for action command detection

• Face recognition and tracking system for mobile robot :

  Developed face tracking system for robotic application

  • Tools & technologies used :VGGface, YOLOv5, PyTorch, openCV

  • Face recognition using the pretrained VGGface model and utilised YOLOv5 models for detecting for person detection

  • Utilised SORT(Simple Online Realtime Tracking) algorithm for tracking overtime

• Physics Informed Neural Networks :

  Currently pursuing my capstone project in PINNs

  • Tools & technologies used :Pytorch, Pandas, Numpy, Python

  • Built a Voice assistant using gemini LLM api for generating responses

  • Wake word detection implemented using picovoice api and text to speech using gTTs library

  • Calculated cosine similarity of sentences using NLTK for action command detection