ATIKA ANJUM

—

—

—

—

—

COMSATS University Islamabad | BS in Artificial Intelligence (3.78/4.00 CGPA) | Expected Graduation: 2026 | Relevant Coursework: Programming Fundamentals, Data Structures, Data Science, Introduction | to AI, Programming Fundamentals of AI, Machine Learning, Natural Language Processing, Deep | Learning, Artificial Neural Network, Computer Vision | TECHNICAL SKILLS | Programming: Python, Java, C++ | Other: HTML/CSS, Flutter | Frameworks: TensorFlow, Pytorch, FastAPI, LangChain, Bootstrap | Skills: Machine Learning, Deep Learning, Natural Language Processing, Computer Vision | Tools: MS Word, Excel, PowerPoint, VS Code, Jupyter Notebook, Google Colab, MongoDB | PROJECT EXPERIENCE | FEbot – Intelligent Agentic Chatbot | Built a bot using Flutter and a RAG-based AI backend to provide guidance, identify |  | user intent, and assist with locating nearby hospitals, police stations, and safe places. | Diabetes-Retinopathy-Detection | Developed a deep learning system for diabetic retinopathy detection using the |  | EfficientNet-B3 CNN, trained on retinal images to classify disease severity across five | stages (0–4). | Voice Emotion Detection | Built a hybrid deep learning model using CNN and BiLSTM networks to recognize |  | emotions in human speech, detecting anger, happiness, sadness, fear, disgust, and neutral | from audio recordings. | Music Genre Classifier | Built a CNN-based music genre classification model using Python for audio feature |  | extraction, applying supervised learning to classify genres. | Vehicle Detection & Depth Analysis | Built a traffic monitoring system using YOLOv5 and OpenCV stereo vision to |  | detect vehicles and estimate their distance from the camera. | Paraphrasing & Multilingual Translation with mBART | Built a fine-tuned mBART model for paraphrasing and multilingual translation using |  | Low-Rank Adaptation (LoRA) for efficient training and Hugging Face Trainer API for | MRPC sentence-pair classification.

—

—