README.md

metadata

title: Object Detection - YOLOV3
emoji: 🚀
colorFrom: green
colorTo: purple
sdk: gradio
sdk_version: 3.40.1
app_file: app.py
pinned: false
license: mit

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Demonstration of Object Detection using YOLOv3. This model is trained using Pytorch Lightning from scratch on PASCAL VOC dataset with 20 classes. The input image resolution used for the training was 416 x 416.

The app also shows the saliency maps of the input images generated using EigenCAM explainability method.

The model supports following classes:

• aeroplane ✈️
• bicycle 🚲
• bird 🐦
• boat 🛥️
• bottle 🍾
• bus 🚌
• car 🚗
• cat 🐈
• chair 🪑
• cow 🐄
• diningtable
• dog 🐕
• horse 🐎
• motorbike 🏍️
• person 👱
• pottedplant
• sheep 🐑
• sofa 🛋️
• train 🚊
• tvmonitor 📺

Steps to Use:

• Upload an image and click submit to generate the detections, the corresponding bounding boxes, and the saliency maps.
• Control the degree of detection threshold and the IOU threshold to filter the predictions