Report

You can download the Report from here / drive

Introduction

In this project, we intend to detect objects that are closer than a threshold using depth estimation and object detection networks.

We used the AdaBins for depth estimation and YOLOX for object detection.

Download Links

• You can download the pretrained models "AdaBins_nyu.pt" and "yolox_m.pt" from here

Quick Start

• Step1. Clone the project.

git clone https://github.com/amindehnavi/DL2021_FinalProject.git

• Step2. Change current directory to "DL2021_FinalProject"

• Step3. Put "AdaBins_nyu.pt" in "./AdaBins/pretrained".

• Step4. Put "yolox_m.pt" in "./YOLOX/pretrained".

• Step5. Install the required packages.

pip3 install -r ./requirements.txt

• Step6. Run "GUI.py".

python ./GUI.py

Cite US

If you use our report in your research, please cite us:

 @report{Sharif Deep Vision,
  title={DL2021 Final Project: Object Detection and Depth Estimation App},
  author={MohammadAmin RamezanDehnavi;ID:99202318, Sina Nabigol;ID:99205361},
  course={Deep Learning, Department of Electrical Engineering, Sharif University},
  year={2021}
}