Shruti k

I am a computer vision and deep learning engineer with hands-on experience building volumetric spectral-spatial models and production-ready ML pipelines.

During a research internship at VNIT Nagpur, I developed hybrid 3D CNN + Vision Transformer models for multi-class land-cover segmentation and created GPU-accelerated, reproducible preprocessing and batching pipelines.

My projects include a hyperspectral/DEM landslide detection system achieving high IoU through rigorous benchmarking and ablation studies, and real-time object detection with YOLOv5 for crop disease identification optimized with mixed precision training.

I combine strong experimentation practices (cross-validation, hyperparameter tuning, statistical analysis) with software engineering skills (Docker basics, Git, serverless architectures) to deliver efficient, reliable ML solutions.