F
Fumin Shen
Researcher at University of Electronic Science and Technology of China
Publications - 244
Citations - 11134
Fumin Shen is an academic researcher from University of Electronic Science and Technology of China. The author has contributed to research in topics: Hash function & Computer science. The author has an hindex of 44, co-authored 221 publications receiving 8028 citations. Previous affiliations of Fumin Shen include University of Adelaide & Nanjing University of Science and Technology.
Papers
More filters
Proceedings ArticleDOI
Movie Fill in the Blank with Adaptive Temporal Attention and Description Update
TL;DR: This paper proposes to use a novel LSTM network called L STM with Linguistic gate (LSTMwL), which exploits adaptive temporal attention for MovieFIB, and demonstrates that this approach outperforms state-of-the-art models for movieFIB.
Book ChapterDOI
Discrete binary hashing towards efficient fashion recommendation
TL;DR: The proposed model jointly learns the intrinsic matching patterns from the matching matrix and the binary representations from the clothing items’ images, where the visual feature of each clothing item is discretized into a fixed-length binary vector.
Book ChapterDOI
Hashing with Inductive Supervised Learning
TL;DR: A new supervised hashing method to generate class-specific hash codes, which uses an inductive process based on the Inductive Manifold Hashing IMH model and leverage supervised information into hash codes generation to address these difficulties and boost the hashing quality.
Posted Content
Highly-Economized Multi-View Binary Compression for Scalable Image Clustering
TL;DR: Wang et al. as discussed by the authors proposed the Highly-Economized Scalable Image Clustering (HSIC) method, which unifies the binary representation learning and efficient binary cluster structure learning into a joint framework.
Journal ArticleDOI
Embedding and predicting the event at early stage
TL;DR: An event early embedding model (EEEM) that can extract social events from noise, find the previous similar events, and predict future dynamics of a new event with very limited information is designed and a denoising approach is derived from the knowledge of signal analysis to eliminate social noise and extract events.