Oct 2025 – Present · 8 mos

Sep 2025 – Oct 2025 · 1 mo

• Audio Deepfake Detection focuses on developing systems that identify whether a voice is genuine or generated by AI. As synthetic voices become increasingly realistic, this research aims to safeguard trust, authenticity, and human identity in the digital world.

Aug 2025 – Sep 2025 · 1 mo

Jun 2025 – Aug 2025 · 2 mos

• Learned core quantum concepts including qubits, superposition and entanglement and how they can complement machine learning for stronger robustness with lean compute and memory
• Explored quantum machine learning frameworks including IBM Qiskit and PennyLane and built hybrid quantum and classical pipelines in PyTorch end to end
• Designed variational circuits and embedding strategies and integrated compact quantum layers with proven classical backbones
• Implemented image classifiers and UNet based segmentation with a quantum bottleneck and clear training loops
• Ran targeted ablations on the placement of quantum components across the encoder bottleneck and decoder to find what works in practice
• Studied robustness under noise and simple adversarial settings and compared quantum aware heads with classical baselines
• Profiled models for clarity and efficiency and maintained reproducible code and experiment logs
• Collaborated closely with Dr. Praful Hambarde and Dr. Amit Shukla and worked with Vinita Shukla on daily debugging reviews and write ups