Table of Content

Recent Advances in Hyper Parameters Optimization, Features Optimization, and Deep Learning for Video Surveillance and Biometric Applications

Submission Deadline: 24 August 2022 (closed)

Guest Editors

Dr. Muhammad Attique Khan, HITEC University Taxila, Pakistan.
Dr. Tallha Akram, COMSATS University Islamabad, Pakistan.
Dr. ShuiHua Wang, University of Leicester, UK.
Dr. Seifedine Kadry, ff: Noroff University College, Norway.


Video surveillance and biometric are important applications in the area of computer vision and machine learning from the couple of years. Human action recognition, gait recognition, person identification, and emotions recognition are important topics of video surveillance and biometric applications. A lot of techniques are introduced in the literature for video surveillance and biometric applications especially action recognition and gait recognition. The traditional techniques used for these applications are not performed well due to large number of datasets. Moreover, the static hyper parameters of deep learning models sometime degrade the recognition accuracy. In addition, the higher amount of data degrades the recognition accuracy and increases the computational time. Therefore, in this research proposal, we will target the advanced techniques for accurate recognition and minimize the computational time.


- Human action recognition using advanced deep learning techniques
- Human gait recognition using advanced deep learning and transfer learning
- Human action recognition using features optimization techniques
- Human emotions recognition using deep learning
- Hyper parameters optimization for action and gait recognition
- Person identification using deep learning
- Object recognition using deep learning
- Person Re-identification

Share Link

WeChat scan