Unsupervised Deep Video Hashing with Balanced Rotation

Authors: Gengshen Wu, Li Liu, Yuchen Guo, Guiguang Ding, Jungong Han, Jialie Shen, Ling Shao

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments have been performed on two realworld datasets and the results demonstrate its superiority, compared to the state-of-the-art video hashing methods.
Researcher Affiliation Collaboration Northumbria University, Newcastle, NE1 8ST, UK Malong Technologies Co., Ltd, Shenzhen, 518000, China Tsinghua University, Beijing, 100084, China Lancaster University, Lancaster, LA1 4YW, UK The University of East Anglia, Norwich, NR4 7TJ, UK
Pseudocode Yes Algorithm 1 Unsupervised Deep Video Hashing
Open Source Code No To bootstrap further developments, the source code will be made publically available. This is a promise for future availability, not current concrete access.
Open Datasets Yes FCVID [Over et al., 2014]: There are 91,223 videos collected from Youtube in the dataset... YFCC [Thomee et al., 2015]: It is the largest public video dataset, containing over 0.8M videos.
Dataset Splits Yes In the training phase, we randomly select 45,611 videos for the train split and the rest is used as the test split. ... 0.6M unlabeled videos form the train split in the unsupervised learning.
Hardware Specification Yes The server configurations are: Intel Core i7-5960X CPU, 64GB RAM and a TITAN X GPU.
Software Dependencies No The algorithm is implemented under the open-source Caffe framework [Jia et al., 2014]. This mentions software by name but lacks specific version numbers for reproducibility.
Experiment Setup Yes In the proposed framework, the number of clusters during the clustering is the only hyper parameter. Figure 2(a) shows the effect on m AP@20 when varying the cluster numbers. ... the performance get saturated when the cluster number reaches 400.