Name: Zengjie Song
Type: User
Company: School of Mathematics and Statistics, Xi’an Jiaotong University
Bio: My research interests include predictive coding, multimodal learning, generative model, and computer vision.
Location: Xi'an, China
Blog: https://zjsong.github.io/
Zengjie Song's Projects
PyTorch Implementation for "Visually Guided Sound Source Separation with Audio-Visual Predictive Coding" (IEEE T-NNLS, 2023)
PyTorch code for "Toward a Controllable Disentanglement Network" (IEEE T-CYB, 2020)
Matlab implementation for "Fast Inference Predictive Coding: A Novel Model for Constructing Deep Neural Networks" (IEEE T-NNLS, 2018)
PyTorch code for "Enhancing Sound Source Localization via False Negative Elimination"
PyTorch code for "Self-Supervised Predictive Learning: A Negative-Free Method for Sound Source Localization in Visual Scenes" (CVPR, 2022)