Y
Yu Liu
Researcher at Katholieke Universiteit Leuven
Publications - 52
Citations - 3216
Yu Liu is an academic researcher from Katholieke Universiteit Leuven. The author has contributed to research in topics: Computer science & Convolutional neural network. The author has an hindex of 14, co-authored 45 publications receiving 2044 citations. Previous affiliations of Yu Liu include Dalian University of Technology & Leiden University.
Papers
More filters
Journal ArticleDOI
Deep learning for visual understanding
TL;DR: The state-of-the-art in deep learning algorithms in computer vision is reviewed by highlighting the contributions and challenges from over 210 recent research papers, and the future trends and challenges in designing and training deep neural networks are summarized.
Journal ArticleDOI
A review of semantic segmentation using deep neural networks
TL;DR: The field of semantic segmentation as pertaining to deep convolutional neural networks is reviewed and comprehensive coverage of the top approaches is provided and the strengths, weaknesses and major challenges are summarized.
Proceedings ArticleDOI
Learning a Recurrent Residual Fusion Network for Multimodal Matching
TL;DR: This work introduces a novel bridge between the modality-specific representations by creating a co-embedding space based on a recurrent residual fusion (RRF) block that adapts the recurrent mechanism to residual learning, so that it can recursively improve feature embeddings while retaining the shared parameters.
Proceedings ArticleDOI
Learning Relaxed Deep Supervision for Better Edge Detection
Yu Liu,Michael S. Lew +1 more
TL;DR: This work builds hierarchical supervisory signals with additional relaxed labels to consider the diversities in deep neural networks to use relaxed deep supervision within convolutional neural networks for edge detection and obtains superior cross-dataset generalization results on NYUD dataset.
Journal ArticleDOI
CNN-RNN: a large-scale hierarchical image classification framework
TL;DR: A high performance network based on the CNN-RNN paradigm is built which outperforms the original CNN and also the current state-of-the-art and is built on top of any CNN architecture which is primarily designed for leaf-level classification.