DaPedestrianDetection

Chair for Computer Aided Medical Procedures & Augmented Reality
Lehrstuhl für Informatikanwendungen in der Medizin & Augmented Reality

Pedestrian Detection and Pose Estimation for Autonomous Driving

Abstract

Autonomous driving systems are right on the corner and one key concern around the development and social acceptance of such systems is safeguarding. In this project, we want to look at the task of pedestrian detection from LiDAR? point clouds and their 3D pose estimation from the RGB camera input. 3D object detection from sparse point cloud data and 3D multi-person pose estimation are two challenging tasks and therefore active research fields in both academia and industry. In this project, we want to integrate the state of the art deep learning methods, train models on synthetic renderings and improve their performance based on safeguarding metrics.

Requirements

An overall understanding of such concepts is needed:

  • 3D computer vision
  • Object detection
  • Human pose estimation
  • Point clouds processing

Hands-on experience with:

  • Deep Learning
  • PyTorch
  • Dockers
  • Git

Literature

  • Qi, Charles R., et al. "Pointnet: Deep learning on point sets for 3d classification and segmentation." CVPR. 2017.
  • Shi, Shaoshuai, Xiaogang Wang, and Hongsheng Li. "Pointrcnn: 3d object proposal generation and detection from point cloud." CVPR. 2019.
  • Zhou, Yin, and Oncel Tuzel. "Voxelnet: End-to-end learning for point cloud based 3d object detection." CVPR. 2018.
  • Pavlakos, Georgios, et al. "Coarse-to-fine volumetric prediction for single-image 3D human pose." CVPR 2017.
  • Mehta, Dushyant, et al. "Xnect: Real-time multi-person 3d human pose estimation with a single rgb camera." arXiv:1907.00837 2019.

Application

For application please contact via e-mail [m.saleh_at_tum.de] with your latest CV and transcript of records.

Students.ProjectForm
Title: 3D Pedestrian Detection and Pose Estimation
Abstract: Autonomous driving systems are right on the corner and one key concern around the development and social acceptance of such systems is safeguarding. In this project, we want to look at the task of pedestrian detection from LiDAR? point clouds and their 3D pose estimation from the RGB camera input. 3D object detection from sparse point cloud data and multiple pedestrian 3D pose estimation are two challenging tasks and therefore active research fields in both academia and industry. In this project, we want to integrate the state of the art deep learning methods, train models on synthetic renderings and improve their performance based on safeguarding KPIs defined.
Student:  
Director: Federico Tombari
Supervisor: Mahdi Saleh
Type: Hiwi
Area: Segmentation, Machine Learning, Computer Vision
Status: open
Start:  
Finish:  
Thesis (optional):  
Picture:  


Edit | Attach | Refresh | Diffs | More | Revision r1.2 - 10 Dec 2019 - 09:46 - MahdiSaleh