Object Detection

This repository contains implementations of state-of-the-art object detection techniques:

• Faster R-CNN, Mask R-CNN, SSD, RetinaNet – implemented in TensorFlow 2
• YOLOv3 – using the Darknet implementation by AlexeyAB

Experiments are run on subsets of popular datasets such as COCO, Pascal VOC, OpenImages, and ImageNet.
The repo also includes scripts for dataset preparation and annotation format conversion (XML → CSV, CSV → TFRecord).

---

Contents

• Faster R-CNN – OpenImages subset
• Mask R-CNN – OpenImages subset
• SSD – Pascal VOC + real-time sign language detection
• RetinaNet – Pascal VOC + custom dataset examples (Scocco_Detector)
• YOLOv3 (Darknet) – ImageNet subset
• Utilities – Annotation conversion (xml_to_csv.py, csv_to_tfrecord.py)

---

Tools & Libraries

• TensorFlow 2, Keras
• Darknet (YOLOv3 by AlexeyAB)
• OpenCV, NumPy, Matplotlib

---

Purpose

This repo is my playground for object detection research and practice.
It brings together multiple detection frameworks, experiments with different datasets, and includes practical utilities for preparing data and training custom models.