In this project we present YOLO v3, a new approach to object detection. YOLO can be expanded as You Only Look Once. The prior work on object detection was carried out as a two-step process. They first generate a potential bounding box in an image and then use a classifier on those boxes. After classification, post-processing is used to refine the bounding boxes, eliminate duplicate detections and rescore the boxes based on the other objects in the scene. These methods are complex and are time-consuming. Also, they are hard to train and optimize because each individual component must be trained separately. YOLO eliminates these problems by carrying out the whole detection process in a single pipeline. As its name suggests, uses a single neural network to predict bounding boxes and class probabilities directly from full images in one evaluation . This project describes the procedure of bird detection and classification using YOLO v3 , inferences from training.
Official Full Name | Student ID (MTech Applicable) | Email (Optional) |
---|---|---|
SHASHANK NIGAM | A0198469A | [email protected] |
RAMDAS KRISHNAKUMAR | A0198518M | [email protected] |
PADMINI RAMESH | A0198465L | [email protected] |
<Github File Link>
: https://github.com/Shashankwer/Yolov3/blob/master/Project_report.docx