The Augmented Vision department at the RPTU Kaiserslautern-Landau offers an opportunity to work with different faculty members on a variety of master projects (8 CP) and seminars (4 CP) each semester. This also includes projects and seminars from the MindGarage group. The topics we offer are related to the applications of deep learning on computer vision tasks, including but not limited to:

• 3D reconstruction
• 3D scene understanding
• Document analysis
• Floor plan analysis
• Image captioning
• Image classification
• Image generation
• Object detection and segmentation
• Video analysis

You can find proceedings of the projects and seminars from the previous semesters on the department website. Instructions for registration can also be found there. The topics from our group are listed below.

Topics in SS 2023

• Object Detection in Videos. (Supervisor: Khurram Azeem Hashmi)

• Deep Learning Model Compression Techniques for High Speed Object Detection. (Supervisor: Talha Uddin Sheikh)

• Optimize a Human Pose Estimation Network using Knowledge Distillation. (Supervisor: Muhammad Saif Ullah Khan)

Past Topics

• (WS 2022/2023) Visualization Recommender. Neha Pradip Jagtap.

• (SS 2022) Graphical Page Object Detection in Document Images using Few-Shot Detection. Keerti Prem Gadde.

• (SS 2022) Object Detection in Videos. Md. Naimur Islam, Gokula Krishna Govindan Ravi.

• (WS 2021/2022) Road Traffic Analysis: Combining Traditional and Deep Learning Approaches. Prasada Hegde.

• (WS 2021/2022) Handwritten Circuit Detection. Vidushi Jain.

• (WS 2021/2022) Exploiting Attention Mechanisms for Video Object Detection. Shishir Muralidhara.

• (SS 2021) Benchmarking for Activity Recognition on Low Resolution. Yu Wei Huang.

• (SS 2021) Model-based Learning from Learned Models in Reinforcement Learning. Muhammad Gul Zain Ali Khan.

• (SS 2021) 3D Point Cloud Shape Completion. Danish Nazir.

• (SS 2021) 3D Reconstruction of Textureless Surfaces. Muhammad Umer Javed, Dardan Haliti, Muhammad Saif Ullah Khan.

• (SS 2021) Unsupervised Skeleton Based Activity Recognition. Jayanth Siddamsetty.

• (SS 2021) Object Detection/Recognition in Floor Plan Images. Shashank Mishra.

• (SS 2021) Impact of Deformable Convolutions on Object Detection Techniques. Lala Shakti Swarup Ray, Prateek Mahadevappa Havanur, Muhammad Khalid, Sankalp Sinha.

• (WS 2020/2021) Object Detection Under Challenging Environment. Muhammad Ahmed.

Prerequisites

Students who want to work on a project or seminar offered by the MindGarage group should have successfully completed the “Very Deep Learning” course and have a good understanding of deep learning concepts.

If you are interested in working with us on a project or seminar, please contact us.