The Research and Postgraduate Centre would like to congratulate Gan Qi Wen, Iftekhar Salam, Wu Yubing, Zhou Luchen, and Dr. Yau Wei Chuen (corresponding author) from the School of Computing and Data Science on publishing in Expert System with Applications, an SCI Q1 journal with the latest impact factor of 7.5.
The research, titled “DSteganoM: Deep steganography for motion capture data”, addresses the unique challenges of using motion capture (mocap) data in steganography. Mocap data is crucial for applications like games, animations, virtual reality, and medical diagnostics. Unlike image, audio, and video steganography, mocap steganography faces challenges due to varying temporal lengths, complex topologies, and non-repetitive movements.
Dr. Yau’s team proposed DSteganoM, a deep neural network (DNN) model designed to encode sequences of skeletal motions into others, utilizing the spatial and temporal redundancy in mocap data. The model employs upsampling and pooling mechanisms to handle diverse data.
Experimental results showed DSteganoM’s superior performance, with strong data reconstruction and resistance to signal perturbations. This research advances digital security applications by exploring deep learning in mocap steganography, opening new possibilities for securely transmitting sensitive information in 3D environments.
The full article can be accessed at https://doi.org/10.1016/j.eswa.2023.121955