AKELLA SAI ADITYA

Detail-oriented Machine Learning and Deep Learning engineer with expertise in real-time computer vision, time-series forecasting, and embedded AI systems. Proficient in Python, TensorFlow/Keras, and OpenCV, with a strong foundation in Electrical and Computer Engineering. Extensive experience with CNNs, RNNs, and YOLO models, driving innovative solutions in ML workflows. Aiming to contribute technical skills and creative problem-solving in ML/Data Science roles.

• Machine Learning
• Computer Vision
• Sports
• Cooking
• Videography
• Travel

Software License Demand Forecasting using LSTM

• Developed a time-series forecasting model using LSTM to predict software license demand.
• Performed EDA, normalization, and rolling window techniques to improve prediction accuracy.
• Applied LSTM to capture patterns and seasonality in historical license usage.
• Helped reduce procurement overhead by forecasting spikes and plateaus in usage trends.

Automatic Detection of Leukemia with Hybrid Features Using Transfer Learning

• Engineered a hybrid ML pipeline using image segmentation + CNN + RNN, improving binary classification accuracy.
• Extracted features from blood smear images and fused them with traditional image metrics for improved detection.

Real-Time Object Detection & People Tracking

• Built an end-to-end object detection pipeline using YOLOv3-v8.
• Integrated object tracking with motion vector analysis to reduce false positives in dynamic video streams.

Real-Time Facial Recognition with OpenCV & Dlib

• Designed a camera-based facial authentication system using Dlib and OpenCV landmarks.
• Stored embeddings in CSV format and matched face identities in real-time with high accuracy.
• 
  

Driver Drowsiness and Distraction Detection System

• Implemented eye aspect ratio and head movement detection using OpenCV to prevent fatigue-related accidents.
• Triggered alerts based on real-time behavioral indicators.