HackSMU elephant-call denoising and bioacoustic ML toolkit for the iMasons AI & ML for Nature challenge

Engineered a computer vision pipeline for real-time object detection in infrared media. Fine-tuned a YOLOv12 model to classify 17 species of wildlife, reaching 88.3% mAP. Built a user-friendly frontend with Streamlit, enabling seamless interaction with the model for both image and video-based detection..